434f466288e74845b8d97cf092a39b5c3ab5682c
100
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 rootandClaude Opus 4.7
|
434f466288 |
matrix: roleNormalize allowlist for non-plural-s tokens (scrum role_gate_v1)
3-lineage scrum review of the role-gate work (commits 7f2f112..0331288)
ran Opus 4.7 / Kimi K2.6 / Qwen3-coder via scripts/scrum_review.sh.
All three flagged the same edge case: the homegrown plural-stripper
in roleNormalize would collapse non-plural-s tokens like "Sales" →
"Sale", "Logistics" → "Logistic", "Operations" → "Operation". In a
staffing domain those are real role names; the silent normalization
would have caused false role-equality matches and re-opened the
cross-role bleed for those clusters.
Fix:
- nonPluralSWords allowlist for known staffing-domain non-plural-s
tokens (sales, logistics, operations, facilities, premises, news,
physics, economics, mathematics, analytics).
- Last-word-only stripping ("Sales Associate" stays whole; only
"Associates" head noun is plural-checked).
- -ss ending check so "Press Operator", "Boss" don't lose their s.
- strings.ToLower + strings.TrimSpace replace the homegrown rune-
loop ASCII normalizer (Opus INFO — minor cleanup, folded in).
Tests:
- TestRoleNormalize_NonPluralS: 18 cases covering the allowlist,
-ss ending, real plurals (Operators → Operator, Boxes → Box),
multi-word real plurals (Forklift Operators → forklift operator),
whitespace/case tolerance.
- TestRoleEqual_NonPluralS: gate-level pairing — proves equal-
shape allowlisted tokens compare equal AND that "Sales" ≠ "Sale"
(the original bug shape).
- Existing TestRoleEqual_PluralAndCase still green (refactor
preserved behavior).
Other scrum findings dispositioned (not actioned):
- Opus WARN on empty-role fail-open semantics: documented
backward-compat behavior; production path closes via opt-in LLM
extractor (real_004).
- Opus INFO on unsynchronized package-global cache map: harness is
single-goroutine; add sync.Mutex when/if it parallelizes.
- Opus INFO on parallel constructor (NewPlaybookEntryWithRole vs
optional arg): API smell only, both forms preserved.
- Kimi 2 BLOCKs (NewPlaybookEntryWithRole missing, ApplyPlaybookBoost
signature breakage): FALSE positives. Pre-push smoke chain green
on
|
||
|
rootandClaude Opus 4.7
|
0331288641 |
playbook_lift: LLM-based role extractor closes shorthand bleed (real_004)
real_003 left a known-weak hole: shorthand-style queries
("{count} {role} {city} {state} ...") have no separator between
role and city, so a regex can't reliably extract — leaving the
cross-role gate disabled when both record AND query are shorthand.
This commit adds a roleExtractor with regex-first + LLM fallback:
- Regex first (fast, deterministic) — handles need + client_first +
looking from real_003b. ~75% of styles, no LLM cost paid.
- LLM fallback when regex returns empty AND model is configured —
Ollama-shape /api/chat with format=json, schema-tight prompt,
temperature 0. ~1-3s on local qwen2.5.
- Per-process cache — paraphrase + rejudge passes reuse the same
query 4× per run; cache prevents 4× LLM cost.
- Off-by-default — opt-in via -llm-role-extract flag (CLI) and
LLM_ROLE_EXTRACT=1 env var (harness wrapper). real_003b shipping
config unchanged unless explicitly enabled.
8 new tests in scripts/playbook_lift/main_test.go:
- TestRoleExtractor_RegexFirst: LLM not called when regex matches
- TestRoleExtractor_LLMFallback: shorthand goes to LLM
- TestRoleExtractor_LLMOffLeavesEmpty: opt-in default preserved
- TestRoleExtractor_Cache: 3 calls = 1 LLM hit
- TestRoleExtractor_NilSafe: nil receiver runs regex only
- TestExtractRoleViaLLM_HTTPError + _BadJSON: failure paths
- TestRoleExtractor_ClosesCrossRoleShorthandBleed: synthetic
witness for the real_003 scenario — both record + query are
shorthand, regex returns "" for both, LLM produces DIFFERENT
role tokens for CNC vs Forklift, so matrix gate's cross-role
rejection (locked separately in
TestInjectPlaybookMisses_RoleGateRejectsCrossRole) fires
correctly. This is the load-bearing verification.
Reality test real_004 ran the same 40-query stress as real_003 with
LLM extraction on. Cross-style same-role boosts fired correctly
across all 4 styles for Loaders + Packers + Shipping Clerk clusters
(including shorthand → other-style transfer). No cross-role bleed
observed. The reality test alone can't be a clean "with vs without"
comparison (HNSW build is non-deterministic across runs, and
real_004 stochastics didn't trigger a shorthand recording at all),
which is why the unit-test witness exists.
Production note (in real_004_findings.md): LLM extraction is for
reality-test coverage of arbitrary query shapes. Production should
extract role at INGEST time (when the inbox parser already runs an
LLM) and pass already-resolved role through requests — same shape
as multi_coord_stress's existing Demand{Role: ...} model. The hot
path should never need the harness extractor's per-query LLM cost.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
3263254f1c |
reality_test real_003: 40-query paraphrase stress + extractor extension
Stress-tests the role gate with 40 queries (10 fill_events rows × 4
styles): need, client_first, looking, shorthand. Each row's role +
client + city stays the same; only the surface phrasing changes.
real_003 (original extractor) confirmed the shorthand-vs-shorthand
failure mode: CNC Operator shorthand recording leaked w-2404 onto
Forklift Operator shorthand query within the same Beacon Freight
Detroit cluster. Both record + query had empty role (extractor
returns "" for shorthand because there's no separator between role
and city), gate disabled, distance check passed, bleed fired.
Fix: extended extractRoleFromNeed to handle client_first
("{client} needs N {role} in...") and looking ("Looking for N
{role} at...") patterns. Shorthand left intentionally unmatched —
"Forklift Operator Detroit" is shape-indistinguishable from
"Forklift" + "Operator Detroit" without an LLM extractor or known-
cities lookup.
real_003b (extended extractor) verifies bleed closed across all 4
styles for this dataset. Forklift Operator queries keep w-2136 (the
cold-pass-correct match) regardless of which style the query came
in. Same-role boosts now fire correctly across styles — a CNC
Operator recording made in `looking` style boosts the CNC need-form
query.
scripts/cutover/gen_real_queries.go: added -styles flag with values
need|client_first|looking|shorthand|all (default need preserves
real_001/002 behavior). Tests/reality/real_coord_queries_v2.txt is
the 40-query stress file.
scripts/playbook_lift/main_test.go: 10 sub-tests lock the four
documented patterns + shorthand limitation + lift-suite-style
queries (no clean role, returns empty as expected).
Aggregate metrics:
- real_003 (original): disc=7, lift=7, boost=14, meanΔ=-0.108
- real_003b (extended): disc=11, lift=10, boost=31, meanΔ=-0.202
The growth reflects more LEGITIMATE same-role same-cluster transfer
firing across styles, not bleed (verified by per-cluster bleed
table — Forklift Operator queries unchanged across all 4 styles).
Known limitation documented in real_003_findings.md: same-cluster,
same-role queries in shorthand still embed close enough that a
shorthand recording could bleed onto a different-role shorthand
query if both record + query strip role. Closing this requires
LLM extraction or known-cities lookup at record + query time.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
997527be4d |
matrix: cross-role playbook gate — closes real_001 bleed (OPEN #1)
real_001 surfaced same-client+city queries bleeding across roles:
Q#2 (Forklift Operator @ Beacon Freight Detroit) recorded e-6193
in the playbook corpus. Q#5 (Pickers same client+city) and Q#10
(CNC Operator same client+city) embedded within 0.13-0.18 cosine of
Q#2's query — well inside the 0.20 inject threshold — so e-6193
injected on both, demoting the cold-pass-correct workers.
Root cause: the inject distance threshold isn't tight enough on
the same-client+city cluster. Cosine collapses queries that share
city + client + count-token + time-token regardless of role. The
existing judge gate is per-injection at record time and doesn't
fire at retrieve time.
Fix: structural role gate in front of both Shape A boost and
Shape B inject. PlaybookEntry gains Role; SearchRequest gains
QueryRole. When both are non-empty and differ under roleEqual's
case+plural normalization, the entry is rejected before BoostFactor
or judge-gate logic runs.
Backward-compat: empty role on either side disables the gate —
preserves behavior for the lift suite's free-form multi-constraint
queries that have no clean single role. Caller-supplied (not
inferred), so existing recordings unaffected.
Wire-through:
- internal/matrix/playbook.go: Role field, NewPlaybookEntryWithRole,
roleEqual helper with plural+case normalization
- internal/matrix/retrieve.go: QueryRole on SearchRequest, threaded
to both ApplyPlaybookBoost + InjectPlaybookMisses
- cmd/matrixd/main.go: role on POST /matrix/playbooks/record + bulk
- scripts/playbook_lift/main.go: extractRoleFromNeed regex pulls
role from "Need N {role}{s} in" queries (the fill_events shape);
free-form queries fall back to empty (gate disabled)
Tests (5 new):
- TestInjectPlaybookMisses_RoleGateRejectsCrossRole: exact Q#10
scenario (distance 0.135, recorded "Forklift Operator", query
"CNC Operator") — locks the bleed at unit level
- TestInjectPlaybookMisses_RoleGateAllowsSameRole: Forklift Operator
recording fires on Forklift Operators query (plural normalization)
- TestInjectPlaybookMisses_RoleGateBackwardCompat: empty Role on
either side = gate disabled, preserves current behavior
- TestApplyPlaybookBoost_RoleGateRejectsCrossRole: Shape A defense
in depth — boost doesn't fire on cross-role even when answer is
in cold top-K
- TestRoleEqual_PluralAndCase: case + -s + -es plural normalization
Verification (real_002, same query set as real_001):
- Q#5 Pickers @ Beacon Freight: e-6193 → e-8499 (no bleed)
- Q#10 CNC Operator @ Beacon Freight: e-6193 → w-2404 (no bleed)
- Discoveries + lifts unchanged at 2 each (same-role lift still fires)
- Mean Δdist tightens from -0.127 to -0.040 (boosts no longer
pulling distances through the floor on cross-role mismatches)
Findings: reports/reality-tests/real_002_findings.md.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
7f2f112e6a |
reality_test real_001: real-shape coordinator queries — surfaces cross-role bleed
First retrieval probe with non-synthetic query distribution. Pulls
N rows from /home/profit/lakehouse/data/datasets/fill_events.parquet
(real-shape demand data) and translates each to the natural language
a coordinator would type: "Need {count} {role}s in {city} {state}
starting at {at} for {client}".
Headline: 8/10 cold-pass top-1 = judge-best on real distribution.
Substrate works on queries it was never trained for. v2-moe + workers
corpus carry the load.
Surfaced finding (the real value of running this): same-client+city
queries cluster, and Shape A's distance boost bleeds across roles
within the cluster. Q#2 (Forklift @ Beacon Freight Detroit) records
e-6193 in the playbook corpus. Q#5 (Pickers same client+city) and
Q#10 (CNC Operator same client+city) inherit e-6193 at warm top-1
even though:
- Neither query has its own recorded playbook.
- Neither warm pass triggers a Shape B inject (boosted=0).
- The roles are different staffing categories.
Q#10 specifically demoted the cold-pass-correct w-3759 (judge rating
4 at rank 0) for a worker who was approved by the judge for a
different role on a different query.
Why the lift suite missed it: synthetic queries use 7 disjoint
scenario buckets (forklift+OSHA+WI / CDL+IL / etc.). Real demand
clusters on (client, city). The cluster doesn't exist in the
synthetic distribution.
Why the judge gate doesn't catch it: the gate (
|
||
|
rootandClaude Opus 4.7
|
5687ec65c2 |
G5 cutover prep: embed parity probe — Rust /ai/embed ↔ Go /v1/embed verified
First concrete cutover artifact: scripts/cutover/embed_parity.sh
brings up Go embedd + gateway alongside the live Rust gateway,
hits both /ai/embed and /v1/embed with the same forced model, and
emits a per-date verdict report under reports/cutover/.
Why embed first: the parity invariant is one math identity (cosine
sim of vectors against same input). Retrieve has thousands of edge
cases. If embed parity holds, all downstream vector consumers
inherit confidence; if it doesn't, we catch it in 30s instead of
after a flip.
Verdict 2026-04-30: 5/5 samples cosine=1.000000 with model forced
to nomic-embed-text (v1). Same with nomic-embed-text-v2-moe (both
Ollamas have it loaded). Math is provably equivalent across the
gateway plumbing.
Drift catalog (reports/cutover/SUMMARY.md):
- URL: Rust /ai/embed vs Go /v1/embed
- Wire: Rust {embeddings, dimensions} (plural) vs Go {vectors,
dimension} (singular). Wire-format adapter is the only real
cutover work for this endpoint.
- L2 norm: Rust unit vectors (~1.0); Go raw Ollama (~20-23). Same
direction (cos=1.0); harmless under cosine-distance HNSW (which
is Go vectord's default), but worth fixing in internal/embed/
before extending to euclidean indexes.
reports/cutover/ now tracked (joined the scrum/ + reality-tests/
exemptions in .gitignore).
Next probe: /v1/matrix/retrieve ↔ Rust /vectors/hybrid for the
real user-facing retrieve path. Embed parity gives that probe a
clean foundation.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
a2fa9a2ce7 |
scripts/scrum_review: pipe diff via temp files — fixes argv overflow on large bundles
`jq --arg` and `curl --data-binary @-` both read stdin/argv-bound buffers. Diffs >~128KB blow past the kernel's argv limit even when piped via stdin (because we still build `body` as a shell variable first, then feed it to curl). Voice-ai full bundle was 156K and hit it. Switch to writing user/system/body to mktemp files, jq reads via --rawfile, curl reads via @file. Same on-the-wire shape, no argv involvement. Cleanup with rm at the end. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
68d9e554b0 |
shared: auto-emit Langfuse trace+span per HTTP request — closes OPEN #2
Adds langfuseMiddleware in internal/shared so every daemon's shared.Run gets free production-traffic trace visibility when LANGFUSE_URL + LANGFUSE_PUBLIC_KEY + LANGFUSE_SECRET_KEY are set. Same env names + file shape as the multi_coord_stress driver, so operators ship one /etc/lakehouse/langfuse.env across the deploy. Wiring is auth-gated: middleware runs INSIDE the RequireAuth group, so 401s from credential-stuffing don't pollute traces. /health is exempt so LB probes don't either. Missing env vars → nil client → middleware is a passthrough no-op (fail-open per ADR-005 5.1). Bundled deploy: - langfuse.env.example template (mode 0640, root:lakehouse) - 11 systemd units gain `EnvironmentFile=-/etc/lakehouse/langfuse.env` (leading - so missing file = OK) - REPLICATION.md bootstrap section documents setup Tests (4): nil passthrough, /health bypass, real-request emission, status-writer wrapping. All green. STATE_OF_PLAY OPEN list: 5 rows → 4 rows. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
5a3364f539 |
matrix: judge-gated Shape B inject — closes lift-suite tail issues
Lift suite run #004 left two unresolved tail issues: - Q6 ("Forklift loader") ↔ Q7 ("Hazmat warehouse, cold storage") swap recordings as warm top-1 because their embeddings are within 0.20 cosine of each other. Distance gate can't tell them apart. - Q9 + Q15 lose paraphrase recovery when qwen2.5 rephrases past the 0.20 threshold. Distance says "drift too far"; sometimes the drift is real (skip), sometimes the paraphrase is still on-domain (don't want to skip). Multi-coord run #008's judge re-rating proved the LLM can distinguish: Q3 crane case landed at distance 0.23 (looks tight) but rating 1 (irrelevant). The judge sees domain mismatch the embedder doesn't. This commit lifts that pattern into the matrix substrate. Shape B inject now optionally routes every candidate through a judge gate before the rank insert lands. Distance + judge BOTH have to approve. internal/matrix/playbook.go: - InjectPlaybookMisses signature gains a query string + an optional InjectGate. nil gate preserves pre-judge-gating behavior (current tests already pass with nil). - New InjectGate interface + InjectGateFunc adapter for tests and non-LLM callers. - Per-candidate gate.Approve(query, hit) call inserted between the dedup and the inject. Rejected candidates skip silently; injected count reflects post-gate decision. internal/matrix/judge.go (new, ~140 lines): - LLMJudgeGate calls an Ollama-shape /api/chat endpoint with the same 1-5 staffing-rubric prompt that worked in multi_coord run #008. fail-closed on HTTP/JSON errors (don't inject if judge can't speak — better miss than wrong-domain). - NewLLMJudgeGate returns nil when URL or Model is empty, matching InjectGate's nil-means-no-judge semantics. internal/matrix/retrieve.go: - SearchRequest gains JudgeURL, JudgeModel, JudgeMinRating fields. Run() builds an LLMJudgeGate when set; passes nil otherwise. Backward compatible — existing callers see no behavior change. Tests: - TestInjectPlaybookMisses_GateRejectsCandidate (rejectAll → 0 injected, even with tight distance) - TestInjectPlaybookMisses_GateApprovesCandidate (approveAll → same as nil-gate behavior) - TestInjectPlaybookMisses_GateSeesCorrectQuery (gate receives CURRENT query + RECORDED query separately so it can score the (current, candidate) pair) - All 5 existing inject tests updated to new signature go test ./internal/matrix → all 8 inject tests pass. go test ./internal/matrix ./internal/shared ./cmd/{matrixd, queryd,pathwayd,observerd} → all green. STATE_OF_PLAY: - OPEN item #1 (judge-gated injection) closed. - DO NOT RELITIGATE adds the substrate-level judge-gate lock. - OPEN list now 5 rows (was 6). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
247e36e687 |
STATE_OF_PLAY: trim OPEN list — 9 rows → 6, ordered by product leverage
Sprint 4 row removed (shipped: |
||
|
rootandClaude Opus 4.7
|
54a05d9311 |
Sprint 4 deployment artifacts: Dockerfile + docker-compose
Parallel deploy target to the systemd units that landed in
|
||
|
rootandClaude Opus 4.7
|
a59ef5b930 |
Sprint 4 deployment artifacts: 11 systemd units + REPLICATION.md + env templates
Builds on ADR-006 to ship the operator-facing bits Sprint 4 was
blocked on. Single-host deploy is now a documented procedure.
deploy/systemd/ (12 files):
- 11 .service units, one per daemon. Each follows the same template:
Type=simple, User=lakehouse, hardening (NoNewPrivileges,
ProtectSystem=strict, ProtectHome, PrivateTmp, ReadWritePaths
scoped to /var/lib/lakehouse + /var/log/lakehouse), JSON to
journald with per-daemon SyslogIdentifier, EnvironmentFile=- on
/etc/lakehouse/auth.env.
- Dependency graph baked in via After=/Requires=:
storaged → standalone (only network-online)
catalogd → Requires storaged
ingestd → Requires storaged + catalogd
queryd → Requires catalogd
matrixd → Requires embedd + vectord
gateway → Wants every other daemon (Wants= not Requires=
so a single upstream restart doesn't cascade-restart
the gateway)
pathwayd / observerd / vectord / embedd / chatd → standalone
- chatd unit reads 4 cloud-provider EnvironmentFile=s
(ollama_cloud / openrouter / opencode / kimi) — each is its own
file so per-provider key rotation doesn't restart the others.
- lakehouse-go.target: convenience aggregator. Operators
systemctl start/stop/enable lakehouse-go.target instead of
managing 11 daemons individually. Per-daemon WantedBy=
this target.
deploy/etc-lakehouse/ (2 templates):
- auth.env.example: AUTH_TOKEN per ADR-006 6.2 + rotation playbook
comments. The committed file is empty — operators copy + fill in.
- secrets-go.toml.example: [s3.primary] template with
REPLACE_ME placeholders. Multi-bucket G2 example commented.
REPLICATION.md (top-level):
- Operator runbook from fresh box → 11 daemons running.
- Prereqs (Go 1.25+, gcc, MinIO, Ollama, optionally Langfuse +
Postgres for Langfuse) with reachability checks.
- Bind ports table (3110–3220, shifted by 10 from Rust legacy).
- Bootstrap: useradd → build → install → config → secrets →
systemd → validation.
- Auth posture matrix (loopback / non-loopback / multi-host / TLS).
- Token rotation procedure inline (ADR-006 Decision 6.5).
- Logs (journalctl), backup paths, troubleshooting matrix.
Validation: systemd-analyze verify passed on all 11 .service files
(only "not executable" warnings, expected since binaries don't live
at /usr/local/bin/lakehouse/ until step 2 of bootstrap runs).
Sprint 4 is now operator-ready. Next: Dockerfile + multi-stage
build for container deploys (separate concern; deploy targets
either systemd OR docker, not both).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
814197cfd3 |
ADR-006: auth posture for non-loopback deploy + token rotation impl
ADR-003 locked the auth substrate; ADR-006 ratifies the operator
playbook + adds two implementation pieces needed for Sprint 4
deployment: env-resolved tokens and dual-token rotation.
Six decisions locked in docs/DECISIONS.md:
- 6.1: Non-loopback bind requires auth.token (mechanical gate at
shared.Run, already implemented; this ratifies it).
- 6.2: Token from env, not TOML. /etc/lakehouse/auth.env (mode 0600)
loaded by systemd EnvironmentFile=. New TokenEnv field on
AuthConfig defaults to "AUTH_TOKEN".
- 6.3: AllowedIPs for inter-service same-trust-domain; Token for
cross-trust-boundary (gateway ↔ external).
- 6.4: /health stays unauthenticated; everything else under
shared.Run is gated. Already implemented; ratified here.
- 6.5: Token rotation is dual-token. New SecondaryTokens []string
on AuthConfig — both primary and any secondary pass auth
during the rotation window. Implemented in this commit.
- 6.6: TLS terminates at the network edge (nginx/Caddy), not
in-process. Daemons stay HTTP-only; internal traffic stays
on private subnets per Decision 6.3.
Implementation:
- internal/shared/config.go: AuthConfig gains TokenEnv +
SecondaryTokens fields. New resolveAuthFromEnv() called by
LoadConfig fills Token from os.Getenv(TokenEnv) when Token is
empty. TokenEnv defaults to "AUTH_TOKEN" so the happy path needs
no TOML config.
- internal/shared/auth.go: RequireAuth pre-encodes Bearer headers
for primary + every secondary token; per-request constant-time
compare walks the slice. Fast path is 1 compare (primary).
Tests:
- TestLoadConfig_AuthTokenFromEnv (3 sub-tests): default env name,
custom token_env, explicit Token wins over env.
- TestRequireAuth_SecondaryTokenAccepted: both primary + secondary
tokens pass during rotation window.
- TestRequireAuth_SecondaryTokensOnly: only-secondary path works
for the case where primary was just promoted-to-empty mid-rotation.
go test ./internal/shared all green; existing auth_test.go
unchanged (constant-time compare path preserved).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
6c93a38093 |
scrum multi_coord_phase3: 4 fixes from cross-lineage review
Cross-lineage scrum on bundle 87cbd10..f971e64 (3,652 lines)
produced 4 actionable findings, all defensive hardening.
1. (Opus WARN) internal/langfuse/client.go:queue
Synchronous Flush at maxBatch threshold blocked the calling
goroutine for the full 5s HTTP timeout when Langfuse hiccupped,
defeating the "best-effort, never blocks calling path" contract
in the package doc. Now fire-and-forget via goroutine.
2. (Opus + Kimi convergent) cmd/observerd/main.go:handleInbox
- Free-form priority string was accepted; "nonsense" passed
through unchecked. Now closed enum: urgent|high|medium|low (+
empty defaults to medium). Tested: TestInbox_RejectsBadPriority.
- No size cap on body, only emptiness check; multi-MB payloads
would bloat observer's ring + JSONL. Now 8 KiB cap returns 413.
Tested: TestInbox_RejectsOversizedBody.
- Subject/sender/tag concatenated into InputSummary without
newline stripping; embedded \n could corrupt JSONL line-based
parsers. New sanitizeInboxField strips \r\n + caps at 256 chars
before interpolation.
3. (Opus INFO) scripts/multi_coord_stress/main.go
Removed dead `must[T]` generic — tracedSearch took over the
fail-fast role for matrix searches, so the helper became unused.
4. (Opus INFO) scripts/multi_coord_stress/main.go:Event
`JudgeRating int` collapsed "judge errored" and "judge said
unrated" both to 0. Changed to *int — nil = errored, 1-5 =
verdict. judgeInboxResult still returns 0 on error; caller
gates on > 0 before assigning.
Dismissed (with rationale):
- Opus WARN ExcludeIDs ordering: verified by code read — filter
applies after sort + before top-K truncation as documented;
no slot waste possible.
- Opus INFO 10 prior-run reports contradict #011: those are
point-in-time snapshots; intentional history.
- Kimi INFO Langfuse error suppression: design intent (best-effort
per package doc).
- Kimi INFO contract schema validation: defer until contract count
grows enough to make hand-edit drift a real risk.
- Kimi INFO paraphrase prompt duplicated across lift + multi_coord:
defer (lift to internal/paraphrase/ when a third consumer appears).
- Qwen HOLD: single-line, no actionable finding.
go test ./cmd/observerd ./internal/langfuse all green; multi_coord
driver builds clean.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
f971e64745 |
g2_smoke: accept nomic-embed-text* family members as default
Pre-push hook caught the regression — the smoke hardcoded
MODEL = "nomic-embed-text" and the bump to nomic-embed-text-v2-moe
in
|
||
|
rootandClaude Opus 4.7
|
db2e57402e |
STATE_OF_PLAY: capture multi-coord stress wave (Phase 1-3 verified)
Anchor was last touched at v4 split-threshold; since then the
multi-coord stress harness landed end-to-end across 11 commits.
Future sessions reading this file need to see the verified state,
not derive from git log.
Major additions:
- New "Multi-coordinator stress test (Phase 1 → 3)" section in
VERIFIED WORKING. 11-row capability table covering per-coord
playbook isolation, diversity metrics, paraphrase handover,
ExcludeIDs swap, fresh-resume two-tier, inbox endpoints, LLM
demand parsing, judge re-rating, Langfuse tracing.
- Substrate-gains list under that section: ExcludeIDs on
SearchRequest, observer.SourceInbox + /observer/inbox,
internal/langfuse client, embedd default bumped to v2-moe,
two-tier fresh_workers index pattern.
- Last-verified bumped to 16:42 CDT on the run #011 anchor.
DO NOT RELITIGATE expanded with five new locks:
1. Boost / inject use SEPARATE thresholds (0.5 / 0.20)
2. Multi-coord product theory is empirically VALIDATED
3. Fresh content uses two-tier indexing (fresh_workers)
4. embedd.default_model = nomic-embed-text-v2-moe (don't downgrade)
5. Inbox flow: parse + search + judge + trace
6. Langfuse Go-side client lives at internal/langfuse/
OPEN list refresh:
- Removed: re-judge metric (shipped as
|
||
|
rootandClaude Opus 4.7
|
5d49967833 |
multi_coord_stress: full Langfuse coverage — every phase + every call
Phase 1c-only tracing (commit
|
||
|
rootandClaude Opus 4.7
|
08a086779b |
multi_coord_stress: fresh_workers two-tier index — fresh-resume now top-1
Runs #003-#009 surfaced the same finding: fresh workers added mid-run to the main 'workers' vectord index (5K items) reliably *absorbed* (HTTP 200) but failed to *surface* in semantic queries even with content-matching prompts. Distances on the verify queries sat at 0.25-0.65 against existing workers; fresh items were beyond top-K. Better embedder (v2-moe) didn't help — distances got TIGHTER on existing items, pushing fresh items further out of reach. Root cause: coder/hnsw incremental adds to a populated graph land in poorly-connected regions and disappear from search traversal. Known property of HNSW post-build adds; not a bug. Fix: two-tier index pattern (canonical NRT search architecture). Fresh content goes to a small "hot" corpus (fresh_workers); main queries include it in the corpora list and merge results. Hot corpus has no recall crowding because it's tiny; periodic batch job (post- G3) merges it into the main index. Implementation: - ensureFreshIndex(hc, gw, name, dim) — idempotent POST /v1/vectors/index. 409 from re-create treated as "already there." - ingestFreshWorker now takes idx parameter so callers can target fresh_workers instead of workers. - multi_coord_stress phase 1b creates fresh_workers index + ingests 3 fresh workers there + searches verifyCorpora=[workers, ethereal_workers, fresh_workers]. Run #010 result: fresh-001 (Senior tower crane rigger NCCCO Chicago) top-1: fresh-001 from fresh_workers, distance 0.143 fresh-002 (Bilingual Spanish/English OSHA trainer Indianapolis) top-1: fresh-002 from fresh_workers, distance 0.146 fresh-003 (FAA Part 107 drone surveyor Chicago) top-1: fresh-003 from fresh_workers, distance 0.129 3/3 fresh workers surface at top-1 — the absorption-but-not- findable issue from runs #003-#009 is closed. All other metrics held: diversity 0.007, determinism 1.000, verbatim handover 4/4, paraphrase handover 4/4, swap Jaccard 0.000, inbox burst all 6 events accepted + traced to Langfuse. This is the final structural fix for the multi-coord stress suite. Phase 3 is feature-complete. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
7e6431e4fd |
langfuse: Go-side client + Phase 1c instrumentation
The Rust side has Langfuse tracing already (gateway/v1/langfuse_trace.rs); this commit lands Go-side parity so the multi-coord stress harness can emit traces visible at http://localhost:3001. internal/langfuse/client.go: - Minimal Trace + Span + Flush API mirroring what the Rust emitter uses. Auth: Basic over public_key:secret_key. - Best-effort posture: errors are slog.Warn'd, never block calling paths. Same fail-open as observerd's persistor (ADR-005 Decision 5.1) — observability is a witness, not a gate. - Events buffered until 50, then auto-flushed; explicit Flush() at process exit. - Each Trace/Span returns its id so callers can build hierarchies. multi_coord_stress driver wiring: - New --langfuse-env flag (default /etc/lakehouse/langfuse.env). Empty / missing / unparseable file → skip tracing with a logged warning; run still proceeds. - Phase 1c (inbox burst) now emits one parent trace + 4 spans per inbox event: 1. observerd.inbox.record (post to /v1/observer/inbox) 2. llm.parse_demand (qwen2.5 → structured fields) 3. matrix.search (parsed query → top-K) 4. llm.judge_top1 (rate top-1 vs original body) Each span carries input/output JSON + start/end times so the Langfuse UI shows a full waterfall per event. Run #009 result: Trace landed: "multi_coord_stress phase 1c inbox burst" Observations attached: 24 (= 6 events × 4 spans) Tags: stress, phase-1c, inbox Browseable at http://localhost:3001 by tag query. Other harness metrics: diversity 0.016, determinism 1.000, verbatim handover 4/4, paraphrase handover 4/4 — all unchanged by the tracing addition (best-effort post in parallel). Phase 1c is the proof-of-concept; future commits can wrap other phases (baseline / merge / handover / split) in traces too. Once that's done, the entire stress run becomes scrubbable in Langfuse without grepping the events JSON. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
ce940f4a14 |
multi_coord_stress: judge re-rates inbox top-1 — recovers honesty signal
Run #007 surfaced a tradeoff: LLM-parsed inbox queries produce much tighter cosine distances (0.05-0.10 in three cases) but lose the "system has no good match" signal that high-distance results give. A coordinator UI showing only distance can't tell wrong-domain matches apart from real ones. Fix: judge re-rates top-1 against the ORIGINAL inbox body (not the LLM-parsed query). Coordinators see both: - distance: how close was retrieval in vector space - rating: does this person actually fit the original ask The pair tells the honest story. Run #008 result on the 6 inbox events: Demand Top-1 Distance Rating Reading ───────────────────────────────────────────────────────────── Forklift Cleveland w-3573 0.29 4 Strong Production Indy e-1764 0.41 3 Adjacent Crane Chicago e-7798 0.23 1 TIGHT BUT WRONG Bilingual safety Indy w-3918 0.05 5 Perfect Drone Chicago e-1058 0.06 5 Perfect (verify e-1058) Warehouse Milwaukee w-460 0.32 4 Strong The crane-Chicago case is the architectural-honesty signal at work: distance 0.23 says "tight match" but the judge says rating 1 reading the original body. A coordinator seeing only distance would ship the wrong worker; coordinator seeing distance+rating sees the disagreement and escalates. Net distribution: 5/6 rated 3+ (acceptable→perfect), 1/6 rated 1 (irrelevant despite tight cosine). The substrate-honesty signal is recovered without losing the LLM-parse quality wins. Cost: 6 extra judge calls (~9s on qwen2.5). Production amortizes when judge runs only on top-1 of high-priority inbox events; the search-cost-vs-quality tradeoff lives in the priority gate. Implementation: - New JudgeRating int field on Event (omitempty so non-judged events stay clean in JSON) - New judgeInboxResult helper, reusing the same prompt structure as playbook_lift's judgeRate. The two could share an internal package if a third judge consumer appears. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
186d209aae |
multi_coord_stress: LLM-parsed inbox demands (qwen2.5)
Replaced the hard-coded DemandQuery on inbox events with an actual
LLM call: each email/SMS body is parsed by qwen2.5 (format=json,
schema-anchored) into structured {role, count, location, certs,
skills, shift}. The driver then composes a query string from those
fields and runs matrix.search.
This is the real-product flow that the Phase 3 stress test was
asking for: real bodies → real LLM parsing → real search. Before
this commit, the DemandQuery was my hand-crafted string, which
made the inbox phase trivial.
Run #007 result vs #006 (same bodies, parser swapped):
All 6 inbox events parsed cleanly — qwen2.5 nailed:
"Need 50 forklift operators in Cleveland OH for Monday day
shift. OSHA-30 + active forklift cert required."
→ {role:"forklift operator", count:50, location:"Cleveland, OH",
certs:["OSHA-30","active forklift cert"], skills:[], shift:"day"}
Other 5 similarly faithful (indy stayed as "indy", count
defaulted to 1 when unspecified, no hallucinated fields).
LLM-parsed queries produced TIGHTER matches than hard-coded:
Demand #006 dist #007 dist Δ
Crane Chicago 0.499 0.093 -82%
Drone Chicago 0.707 0.073 -90%
Bilingual safety 0.240 0.048 -80%
Forklift Cleveland 0.330 0.273 -17%
Production Indy 0.260 0.399 +53%
Warehouse Milwaukee 0.458 0.420 -8%
Three matches landed at distance < 0.10 — verbatim-replay-tight
territory. Structured queries embed sharper than conversational
hand-crafted strings.
Other metrics unchanged: diversity 0.000, determinism 1.000,
verbatim handover 4/4, paraphrase handover 4/4.
Tradeoff worth flagging: the drone-Chicago case dropped from
distance 0.71 (clear "we don't have one") to 0.07 (confident match
returned). The OOD honesty signal weakens when LLM-parsed structure
makes any closest-neighbor look tight. Future Phase 4 work: judge
re-rates the top match before surfacing, so coordinators see "your
demand was for X but the closest match scored 2/5" rather than just
the worker ID + distance.
Substrate cost: +6 LLM calls per inbox burst (~9s on qwen2.5).
Production would amortize via a small dedicated parser model.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
e7fc63b216 |
observerd: /observer/inbox + multi-coord stress phase 1c (priority-ordered events)
Phase 3 ask: real-world inbox-style event injection during the stress
test. Coordinators in production receive emails + SMS that trigger
contract responses; the substrate has to RECORD these signals AND
react with a search using the embedded demand. This commit lands the
endpoint and exercises it end-to-end in the stress harness.
observerd surface:
- New POST /observer/inbox route — accepts {type, sender, subject,
body, priority, tag} and records as ObservedOp with
Source=SourceInbox. Type must be email|sms; body required;
priority defaults to medium. The handler ONLY records — downstream
triggers (search, ingest, etc.) are the caller's concern, recorded
separately. Keeps the witness role pure.
- New observer.SourceInbox = "inbox" alongside SourceMCP /
SourceScenario / SourceWorkflow.
- Three contract tests on the new route (happy path / bad type / empty
body), router-mount test extended, all green.
Stress harness phase 1c (Hour 9):
- 6 inbox events fire in priority order (urgent → high → medium):
2 urgent emails (forklift Cleveland, production Indianapolis)
1 high email (crane Chicago)
1 high sms (bilingual safety Indianapolis)
1 medium sms (drone Chicago)
1 medium email (warehouse Milwaukee FYI)
- Each event:
1. POSTs to /v1/observer/inbox (recorded by observerd)
2. Triggers matrix.search using a parsed demand (the demand
extraction is hard-coded for now; production needs a small
LLM to parse from body)
3. Captures both as events in the run JSON
Run #006 result (with v2-moe embedder + all phases including inbox):
Diversity:
Same-role-across-contracts Jaccard = 0.000 (n=9)
Different-roles-same-contract Jaccard = 0.046 (n=18)
Determinism: 1.000
Verbatim handover: 4/4 (100%)
Paraphrase handover: 4/4 (100%)
Inbox burst:
6/6 events accepted by observerd (200 status, all recorded)
6/6 triggered searches produced distinct top-1 worker IDs
distance distribution: 0.24 (Indy production) → 0.71 (Chicago
drone surveyor — honest stretch since drones aren't in the
5K-worker corpus, system surfaces closest neighbor at high
distance rather than fabricating)
The drone-Chicago case is the architectural-honesty signal: when
the demand asks for a specialist NOT in the roster, the system
returns the closest semantic neighbor with a distance that flags
"this is a stretch." Coordinators reading distances see "we don't
have a great match here" rather than a confident wrong answer.
Total events captured: 67 (was 61 pre-inbox).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
4da32ad102 |
embedd: bump default to nomic-embed-text-v2-moe (475M MoE, 768d drop-in)
Local Ollama has three embedding models loaded: nomic-embed-text:latest 137M 768d (previous default) nomic-embed-text-v2-moe:latest 475M 768d (this commit's default) qwen3-embedding:latest 7.6B 4096d (would require dim change) v2-moe is a drop-in upgrade — same 768 dim, 3.5× more params, MoE architecture. Workers index doesn't need rebuilding, just future ingests embed with the stronger model. Run #005 result on the multi-coord stress suite: Diversity (same-role-across-contracts): 0.080 → 0.000 (n=9) → MoE is more discriminating: zero worker overlap across Milwaukee / Indianapolis / Chicago for shared role names. The geo + cert + skill context fully separates worker pools. Different-roles-same-contract: 0.013 → 0.036 (still ~96% diff) Determinism: 1.000 (unchanged) Verbatim handover: 4/4 (100%) Paraphrase handover: 4/4 (100%) 200-worker swap: Jaccard 0.000 (unchanged — still perfect) Fresh-resume verify: STILL doesn't surface fresh workers in top-8. With v2-moe, distances increased (top-1 = 0.43–0.65 vs v1's 0.25–0.39) — the embedder is MORE discriminating, but the fresh worker's vector still doesn't outrank the 8th-best existing worker. Now suspect of being an HNSW post-build add issue (coder/hnsw incremental adds can land in hard-to-reach graph regions, not an embedder problem). Better embedder didn't fix it; needs a different strategy: full index rebuild after fresh adds, or explicit playbook-layer score boost for fresh workers, or hybrid (keyword + semantic) retrieval. Phase 3 investigation. Cost: ingest is ~5× slower (workers 20s→100s; ethereal 35s→112s). Acceptable for the quality jump on diversity. Real production with incremental ingest won't pay this once-per-deploy. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
84a32f0d29 |
multi-coord stress Phase 2: ExcludeIDs + fresh-resume + 200-worker swap
Three Phase 2 additions land in this commit: 1. matrix.SearchRequest gains ExcludeIDs ([]string) — filters specific worker IDs out of results post-retrieval, AND skips them at the playbook boost+inject step (so excluded answers can't sneak back via Shape B). Real-world driver: coordinator placed N workers, client asks for replacements, system needs alternatives, not the same N. Threaded through retrieve.go after merge but before metadata filter so excluded IDs don't waste post-filter top-K slots. 2. New harness phase 2b: 200-worker swap simulation. Captures the top-K from alpha's warehouse query, then re-issues with exclude_ids=<placed>. Result Jaccard(orig, swap) measures whether the substrate finds genuine alternatives. 3. New harness phase 1b: fresh-resume mid-run injection. Three new workers ingested via /v1/embed + /v1/vectors/index/workers/add, then verified findable via semantic queries matching resume content. Plus Hour labels on every event (operational narrative: 0/6/12/18/ 24/30/36/42/48) and a refactor of captureEvent to take hour as a param. Run #003 + #004 results (5K workers + 10K ethereal): Diversity (#004): Same-role-across-contracts Jaccard = 0.080 (n=9) Different-roles-same-contract Jaccard = 0.013 (n=18) Determinism: 1.000 (#004 unchanged) Verbatim handover: 4/4 = 100% Paraphrase handover: 4/4 = 100% Phase 2b — 200-worker swap (Jaccard 0.000): 8 originally-placed workers fully replaced by 8 alternatives. ExcludeIDs substrate change works end-to-end — boost AND inject both honor the exclusion, so excluded workers don't return via the playbook either. Phase 1b — fresh-resume injection: REAL PRODUCT FINDING. Substrate ABSORPTION is fine — 3 /v1/vectors/index/workers/add calls at 200 status, 3 vectors persisted. But none of the 3 fresh workers surfaced in top-8 even with semantic queries matching their resume content (e.g. "Senior tower crane rigger NCCCO Chicago" vs fresh-001's resume "Senior rigger with 12 years tower-crane signaling..." NCCCO + Chicago). Top-1 came from existing workers at distance ~0.25; fresh workers' distances must be > 0.25, pushing them past rank 8. Cause: dense retrieval at 5000+ workers means many existing profiles cluster near any specific query in cosine space; nomic-embed-text-v2 (137M) introduces enough noise that a fresh worker doesn't reliably outrank them just because the text content overlaps. Workarounds (Phase 3 work): (a) hybrid retrieval (keyword + semantic), (b) playbook-layer score boost for fresh adds, (c) larger embedder. Documented in run #004 report. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
0fa42a0cc3 |
multi-coord stress Phase 1.5: shared-role contracts + paraphrase handover
Phase 1 had two known gaps: (1) the 3 contracts had zero shared role names, so same-role-across-contracts Jaccard was vacuous (n=0); (2) the verbatim handover at 100% was the trivial case, not the hard learning test (paraphrased queries against another coord's playbook). Both fixed in this commit. Contract redesign — all 3 contracts now share warehouse worker / admin assistant / heavy equipment operator roles, plus a unique specialist per contract (industrial electrician / bilingual safety coord / drone surveyor — the "specialist not on the standard roster" case from J's spec). Counts and skill mixes vary per region. New driver phase 4b — paraphrase handover. Bob runs qwen2.5-paraphrased versions of Alice's contract queries against Alice's playbook namespace. Tests whether institutional memory propagates across coordinators AND across natural wording variation that Bob would introduce when running Alice's contract. Run #002 result (5K workers + 10K ethereal_workers, 4 demand × 3 coords + paraphrase handover): Diversity (the question J asked: locking or cycling?): Same-role-across-contracts Jaccard = 0.119 (n=9) → 88% of workers DIFFER across regions for the same role name. Milwaukee warehouse vs Indianapolis warehouse vs Chicago warehouse pull mostly distinct top-K from the same population. The system locks into geo+cert+skill context, not cycling. Different-roles-same-contract Jaccard = 0.004 (n=18) → role-specific retrieval works (unchanged from Phase 1). Determinism: Jaccard = 1.000 (n=12) — unchanged. Learning: Verbatim handover 4/4 = 100% (trivial case, expected) Paraphrase handover 4/4 = 100% (HARD case — passes!) Of those 4 paraphrase recoveries: - 2 used boost (Alice's recording was already in Bob's paraphrase top-K; ApplyPlaybookBoost re-ranked to top-1) - 2 used Shape B inject (recording wasn't in Bob's paraphrase top-K; InjectPlaybookMisses brought it in) The boost/inject mix is healthy — both paths are used and both produce correct top-1s. Multi-coord institutional memory propagation is empirically working under wording variation. Sample warehouse worker top-1s across contracts (proves diversity): alice / Milwaukee → w-713 bob / Indianapolis → e-8447 carol / Chicago → e-7145 Three different workers from the same 15K-person population, selected on geo+cert+skill context. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
61c7b55e48 |
multi-coord stress harness — Phase 1 of 48-hour mock
Three coordinators (alice / bob / carol) with three contracts
(Milwaukee distribution / Indianapolis manufacturing / Chicago
construction). 7-phase scenario runner: baseline → surge → merge →
handover → split → reissue → analysis. Each coord has a separate
playbook namespace (playbook_{name}) so institutional memory stays
isolated by default but transferable on demand.
Phase 1 deliberately skips the 48-hour clock, email/SMS endpoints,
and Langfuse tracing — those are Phase 2/3.
Run #001 (52 events, 4 queries × 3 coords × 2 demand flavors):
Diversity:
Different-roles-same-contract Jaccard = 0.004 (n=18)
→ role-specific retrieval is working perfectly. Different
roles within one contract pull totally different worker
pools. System is NOT cycling; locks into per-role retrieval.
Same-role-across-contracts Jaccard = N/A (n=0)
→ TEST-DESIGN ISSUE: the 3 contracts use distinct role names
per industry (warehouse worker / production worker / general
laborer), so no exact-name overlaps exist. Phase 2 should
either share at least one role across contracts OR add a
skill-based diversity metric.
Determinism: Jaccard = 1.000 (n=12)
→ HNSW + Ollama retrieval is fully deterministic on identical
query text. coder/hnsw + nomic-embed-text are stable.
Learning: handover hit rate = 4/4 = 100%
→ Bob inherits Alice's recordings perfectly when bob runs
identical queries with alice's playbook namespace. CAVEAT:
this tests the trivial verbatim case, not paraphrase handover.
The harder test (bob runs paraphrased queries with alice's
playbook) is Phase 2 work.
Per-event capture in JSON: every matrix.search response is logged
with phase / coordinator / contract / role / query / top-K IDs +
distances + per-corpus counts + boosted/injected counts. Reviewable
via:
jq '.events[] | select(.phase == "merge")'
jq '.events[] | select(.coordinator == "alice")'
jq '.events[] | select(.role == "warehouse worker")'
Notable finding from per-event: carol's "general laborer" and "crane
operator" queries both surface w-1009 as top-1, with crane operator
at distance 0.098 (very tight) and general laborer at 0.297. The
system found a worker who legitimately covers both roles — realistic
for small construction crews.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
b13b5cd7a1 |
playbook_lift v4 metric: warm-top-1 re-judge — quality lift +24%/-14%
The rank-based "lift" metric (warm-top-1 == cold-judge-best) doesn't distinguish "Shape B surfaced a strictly-better answer" from "Shape B shuffled ranks but quality is unchanged" from "Shape B replaced a good answer with a wrong one." This commit adds Pass 4: judge warm top-1 with the same prompt as cold ratings, then bucket the comparison. Implementation: - New --with-rejudge driver flag (default off). - New WITH_REJUDGE harness env (default 1, on for prod runs). - queryRun gains WarmTop1Metadata (cached during Pass 2 for the rejudge call) + WarmTop1Rating *int (nil-distinguishable; nil = no rejudge, 0..5 = rating). - summary gains RejudgeAttempted, QualityLifted, QualityNeutral, QualityRegressed (counts of warm-rating > / == / < cold-rating). - Markdown headline gains a Quality block when rejudge ran. - ~21 extra judge calls (~30s on qwen2.5). Run #005 result (split inject threshold 0.20 + paraphrase + rejudge): Quality lifted 5 / 21 (24%) — 3× +2 rating, 2× +1 rating Quality neutral 13 / 21 (62%) — includes OOD queries holding 1 Quality regressed 3 / 21 (14%) Net rating delta +3 across 21 queries (+0.14 average) The 5 lifts were all rating-2 cold replaced with rating-3 or rating-4 warm — Shape B took mediocre matches and substituted substantively better ones. The 3 regressions were small (-1, -1, -3). Q11 is the cautionary tale: cold top-1 "production line worker" (rating 4) got replaced by Q1's recorded "forklift OSHA-30 operator" e-5729 (rating 1). Adjacent-domain cross-pollination — production worker and forklift operator embed within 0.20 cosine because both are warehouse-adjacent staffing queries, even though the judge correctly distinguishes them. The split-threshold defense (0.5 boost / 0.20 inject) catches OOD cross-pollination (Q19/Q20/Q21 all stayed neutral at rating 1) but not adjacent-domain cross-pollination. Net product verdict: working, net-positive on quality, but the worst case (Q11 4→1) is customer-visible and warrants a tighter inject threshold OR an additional gate beyond cosine distance. Filed in STATE_OF_PLAY OPEN as a follow-up. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
87cbd10090 |
STATE_OF_PLAY: v4 split-threshold result + adjacent-query observation
- Reality test table extends from #001-#003 to #001-#004; v4 row marked as "the honest configuration" because OOD cross-pollination is gone. - Shape B section gains the split-threshold rationale (boost safe at loose, inject structurally riskier so tighter). - Verbatim drop framing rewritten — v3→v4 is configuration evolution, not regression. - OPEN: closed "Shape B cap/decay" + the conditional Q15 boost-math item (Shape B + split threshold addressed both). Replaced with two finer-grained follow-ups: adjacent-query Q6↔Q7 swap (might be correct, verify with v4 re-judge metric) and liberal-paraphrase recovery loss (Q9/Q15 missed because qwen2.5 drifted >0.20). - RECENT VERIFIED WAVE adds |
||
|
rootandClaude Opus 4.7
|
67d1957b87 |
matrix: split boost / inject thresholds — kills Shape B cross-pollination
Run #003 surfaced rampant cross-pollination: Q2's "OSHA-30 forklift Wisconsin" recording (w-4435) became warm top-1 for Q19 (dental hygienist), Q20 (RN), Q21 (software engineer), and 6 other unrelated staffing queries. Cause: InjectPlaybookMisses inherited the same DefaultPlaybookMaxDistance (0.5) as the boost path, but inject is structurally riskier than boost — boost only re-ranks results that already retrieved on their own merits, while inject FORCES a result into top-K, so a loose match cross-pollinates wrong-domain answers. Empirical motivation from v3: Implied playbook hit distances for cross-pollinated cases: 0.20-0.46 Implied distances for the 6/6 paraphrase recoveries: 0.23-0.30 Threshold of 0.20 should keep most paraphrases, kill the OOD bleed. Implementation: - New DefaultPlaybookMaxInjectDistance = 0.20 in playbook.go. - New PlaybookMaxInjectDistance field on SearchRequest (override). - InjectPlaybookMisses signature gains maxInjectDist param; hits whose Distance exceeds it are skipped (boost path may still re-rank them). - TestInjectPlaybookMisses_RespectsInjectThreshold locks the contract with one tight + one loose hit, asserting only the tight one injects. - Existing tests pass explicit threshold (0 = default for tight tests, 0.5 for the dedupe test which uses 0.30 hits). Run #004 result on identical queries with the split threshold: Verbatim discovery 8 (vs v3's 6 — judge variance, separate) Verbatim lift 6 / 8 (75%) Paraphrase top-1 6 / 8 (75%) Paraphrase any-rank in K 6 / 8 OOD queries Q19/Q20/Q21 ALL show warm top-1 = cold top-1 (no injection) — cross-pollination eliminated where it was wrong-direction. Mean Δ top-1 distance dropped from -0.164 (v3, distorted) to -0.071 (v4, comparable to v1's -0.053). Two paraphrases missed in v4 (Q9, Q15) were ones where qwen2.5 rephrased liberally enough to drift past 0.20 — Q9: "Inventory specialist..." → "Individual needed for inventory management..." and Q15: "Engaged warehouse associate..." → "Warehouse associate currently engaged with a robust history...". The system correctly refusing to inject when it's not confident is the right product behavior; the boost path still re-ranks recorded answers when they appear in regular retrieval. The Q6 ↔ Q7 cross-pollination ("Forklift-certified loader" ↔ "Hazmat warehouse worker") is legitimate — these are genuinely similar staffing queries and the judge ranks both directions as plausible. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
94fc3b67ec |
STATE_OF_PLAY: capture v3 reality test + Shape B + cross-pollination
- Reality test section now spans v1/v2/v3 across one table — the product story (boost-only verbatim → paraphrase gap → Shape B closes the gap) is legible without reading the reports. - Verbatim-lift drop v1→v3 (7→2) explicitly framed as cross-pollination, NOT regression — and filed as v4 re-judge metric in OPEN. - "DO NOT RELITIGATE" gains: Shape B is the stance now (don't revert to boost-only); local_judge stays on qwen2.5 (don't bump to qwen3.5 for cleanliness — vision-SSM cost geometry). - OPEN list: removed the now-closed paraphrase v2 row + the boost-math Q15 row (Shape B may have addressed it; flagged for verify after v4). Added v4 re-judge metric and Shape B injection cap/decay design call. - RECENT VERIFIED WAVE adds the four new commits past |
||
|
rootandClaude Opus 4.7
|
154a72ea5e |
matrix: Shape B — inject playbook misses + 6/6 paraphrase recovery
The v0 boost-only stance documented in internal/matrix/playbook.go:22-27
("the boost only re-ranks results that ALREADY surfaced from the regular
retrieval") couldn't promote recorded answers that dropped out of a
paraphrase's top-K. playbook_lift_002 surfaced exactly that gap: 0/2
paraphrase recoveries because the recorded answers weren't in regular
retrieval at all (rank=-1).
Shape B: when warm-pass retrieval doesn't surface a playbook hit's
answer, inject a synthetic Result for it directly. Distance =
playbook_hit_distance × BoostFactor — same formula as the boost path so
injections land in comparable distance space. Caller re-sorts +
truncates after both boost and inject have run.
Result on playbook_lift_003 (Shape B + paraphrase pass):
Verbatim discovery 6
Verbatim lift 2 / 6
**Paraphrase top-1** **6 / 6**
Paraphrase any-rank in K 6 / 6
Mean Δ top-1 distance -0.1637 (warm closer than cold)
Every paraphrase the judge generated landed the v1-recorded answer at
top-1 of the new query's results. The learning property holds — cosine
on embed(paraphrase) finds the recorded query's vector within
DefaultPlaybookMaxDistance (0.5), and Shape B injects the answer.
Verbatim lift dropped from v1's 7/8 because Shape B cross-pollinates
recorded answers across queries. w-4435 (Q2's recording) appears as
warm top-1 for several other queries because their embeddings are
within the playbook hit threshold of "OSHA-30 forklift Wisconsin." This
is a feature, not a bug — the matrix layer's purpose is to share
knowledge across queries — but the lift metric only counts "warm top-1
== cold judge best," so cross-pollinated lifts don't register. A v3
metric would re-judge warm pass to measure true judge improvement.
Tests:
- TestInjectPlaybookMisses_AddsMissingAnswers — primary claim
- TestInjectPlaybookMisses_SkipsAnswersAlreadyPresent — no double-inject
- TestInjectPlaybookMisses_DedupesPerAnswer — multi-hit same answer
- TestInjectPlaybookMisses_EmptyHits — fast-path no-op
Driver fix: ParaphraseRecordedRank int → *int. The `omitempty` int
silently dropped rank=0 (top-1, the WANTED value) from JSON, making the
v003 report show "null" instead of "0" for every successful recovery.
Pointer keeps nil/rank-0 distinguishable.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
e9822f025d |
playbook_lift v2: paraphrase pass + run #002 finds boost-only limit
Adds an opt-in Pass 3 to the lift driver: for each query whose Pass 1 recorded a playbook, ask the judge to rephrase the query, then re-query with playbook=true and check whether the recorded answer surfaces in top-K. This is the test the v1 report's caveat #3 explicitly flagged as the actual learning-property gate (not the cheap verbatim case). Implementation: - New flag --with-paraphrase on the driver (default off). - New WITH_PARAPHRASE env in the harness (default 1, on for prod runs). - New paraphrase_* fields on queryRun + summary, // 0 fallback in jq so re-rendering verbatim-only evidence stays clean. - generateParaphrase() calls the same judge model with format=json and a tight schema; temperature=0.5 for variance without domain drift. - Markdown report adds a paraphrase per-query table (only when the pass ran) and an honesty caveat about judge-also-rephrases coupling. Run #002 result (reports/reality-tests/playbook_lift_002.{json,md}): Verbatim lift 2/2 (100% — Q7 + Q13, both stable from v1) Paraphrase top-1 0/2 Paraphrase any-rank in K 0/2 Both paraphrases dropped the recorded answer OUT of top-K entirely (rank=-1). This isn't a paraphrase-quality problem — qwen2.5's outputs preserved intent ("Hazmat-certified warehouse worker comfortable with cold storage" → "Warehouse worker with Hazmat certification and experience in cold storage"). It's the v0 boost-only stance documented in internal/matrix/playbook.go:22-27: the boost only re-ranks results that ALREADY surfaced from regular retrieval. If paraphrase's cosine retrieval doesn't include the recorded answer in top-K, no boost can promote it. The "Shape B" upgrade mentioned in the playbook.go comment — inject playbook hits directly even when they weren't in the top-K — is what would close this gap. The reality test surfaced exactly the gap the docs warned about. Worth filing as the next product gate. Run-to-run variance also visible: v1 had 8 discoveries, v2 had 2. HNSW insertion order + judge variance both contribute. Stability of Q7 and Q13 across both runs (lifted in v1 AND v2) is the most reliable signal in the dataset. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
9ce067bd9d |
observerd: test that locks ADR-005 Decision 5.3
TestWorkflowRun_AllProvenanceRecordedPostRun proves that
handleWorkflowRun records ObservedOps only AFTER runner.Run returns,
not interleaved with node execution.
The test pauses inside a node via a controlled channel, samples
observer.Store mid-run (must be 0), unblocks, then samples again
(must be N). If a future commit adds per-node streaming (e.g.
runner.NodeHook firing as each node finishes), n1's record would
appear before the unblock and the first assertion fires.
This is intentional test-as-spec lock. Closing the streaming gap is
deferred per the ADR ("acceptable for short workflows; streaming
callback is the right shape when workflows get longer") — but if
someone later adds the streaming callback without updating the ADR,
this test catches it in `go test` instead of leaving the doc and
code drifted.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
2c71d1c637 |
ADR-005: observer fail-safe semantics
Closes the OPEN item from STATE_OF_PLAY. Required because observerd is
now on the prod-realistic data path via the lift harness boot (
|
||
|
rootandClaude Opus 4.7
|
6c02c905c8 |
scrum lift_001: 4 fixes from cross-lineage review
Cross-lineage scrum on
|
||
|
rootandClaude Opus 4.7
|
b2e45f7f26 |
playbook_lift: harness expansion + reality test #001 (7/8 lift, 87.5%)
The 5-loop substrate's load-bearing gate is verified — playbook +
matrix indexer give the results we're looking for. Per the report's
rubric, lift ≥ 50% of discoveries means matrix is doing real work;
7/8 = 87.5% blew through that.
Harness was structurally hiding bugs behind a 5-daemon stripped boot.
Expanding to the full 10-daemon prod stack surfaced 7 fixes in cascade:
1. driver→matrixd: {"query": ...} → {"query_text": ...} field name
2. harness temp toml missing [s3] → wrong default bucket → catalogd
rehydrate 500 on first call
3. harness→queryd SQL probe: {"q": ...} → {"sql": ...} field name
4. expand boot from 5 → 10 daemons in dep-ordered launch
5. add SQL surface probe (3-row CSV ingest → COUNT(*)=3 assertion)
6. candidates corpus was synthetic SWE-tech (Swift/iOS, Scala/Spark) —
wrong domain for staffing queries; replaced with ethereal_workers
(10K rows, real staffing schema, "e-" id prefix to avoid collision
with workers' "w-"). staffing_workers driver gains -index-name +
-id-prefix flags so the same binary serves both corpora
7. local_judge qwen3.5:latest is a vision-SSM 256K-ctx build running
~30s per judge call against the lift loop; reverted to
qwen2.5:latest (~1s/call, 30× faster, held lift theory)
Each contract drift (1, 3) is now locked into a cmd/<bin>/main_test.go
so future drift fires in `go test`, not in a reality run. R-005 closed:
- cmd/matrixd/main_test.go (new) — playbook record drift detector +
score bounds + 6 routes mounted
- cmd/queryd/main_test.go — wrong-field-name drift detector
- cmd/pathwayd/main_test.go (new) — 9 routes + add round-trip + retire
- cmd/observerd/main_test.go (new) — 4 routes + invalid-op + unknown-mode
`go test ./cmd/{matrixd,queryd,pathwayd,observerd}` all green.
Reality test results (reports/reality-tests/playbook_lift_001.{json,md}):
Queries 21 (staffing-domain, 7 categories)
Discoveries 8 (judge ≠ cosine top-1)
Lifts 7/8 (87.5%)
Boosts triggered 9
Mean Δ distance -0.053 (warm closer than cold)
OOD honesty dental/RN/SWE rated 1, no fake matches
Cross-corpus boosts confirmed (e- ↔ w- swaps in lifts)
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
740eb0d00c |
scrum_review: switch curl to stdin so large diffs don't blow argv
Phase 4-bundle review (128KB diff) hit "Argument list too long" when curl --data was passed the body as a literal arg. Pipe via stdin with --data-binary @- instead. Lifts the practical bundle size from ~30KB to whatever fits in process memory. Caught while running the harness scrum on golangLAKEHOUSE today — the bigger Phase A+B harness diff (4566 lines) tripped it. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
511083ae40 |
docs: SPEC §3.9 (chatd) + §3.10 (local-review-harness sibling)
- SPEC §1 component table: add chatd row marked DONE; replaces Rust gateway's v1::ollama_cloud / openrouter / opencode adapters + the aibridge crate. - SPEC §3.9 — chatd shipped: 5-provider routing (ollama, ollama_cloud, openrouter, opencode, kimi) by model-name prefix or :cloud suffix. Captures the Anthropic 4.7 temperature-deprecation quirk + the local-Ollama think=false default that the playbook_lift judge needed. Mentions scrum_review.sh as the reusable cross-lineage vehicle eating chatd's own /v1/chat. - SPEC §3.10 — local-review-harness sibling tool: separate repo at git.agentview.dev/profit/local-review-harness, MVP shipped today. Documents the cross-pollination plan for when both substrates stabilize (chatd as the harness's LLM backend; harness findings as Lakehouse pathway-memory drift signal; .memory/known-risks as a matrix corpus). Explicit "don't re-port" so future Claudes don't try to absorb the harness into Lakehouse. - STATE_OF_PLAY.md: SIBLING TOOLS section with 1-line summary + pointer to SPEC §3.10. No code changes. just verify still PASS — touched only docs. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
c5c31b6ca6 |
docs: STATE_OF_PLAY.md — Go-side truth anchor (mirrors Rust convention)
Adds the "verified working RIGHT NOW / DO NOT RELITIGATE / OPEN" anchor at the repo root, mirroring /home/profit/lakehouse/STATE_OF_PLAY.md. Memory files (project_golang_lakehouse.md) supplement; this file is the verified-truth pointer. Sections: - VERIFIED WORKING: 13 cmd binaries + 18 smokes + 5 matrix components + Mem0 pathway + observerd + workflow runner + chatd 5-provider dispatcher + model tier registry. just verify PASS in 31s. - DO NOT RELITIGATE: 4 ratified ADRs (DECISIONS.md ADR-001..004) + today's scrum dispositions (B-1..B-4 fixed, FP-A1/A2/C1 dismissed) + session frame items (Rust legacy is maintenance-only, etc.). - OPEN: reality test held on J's queries, 3 daemon main_test.go gap, Sprint 4 deployment, ADR-005 observer fail-safe, ADR-006 auth posture. - RECENT WAVE: 6-commit table 05273ac..e4ee002 documenting today's 4 phases + scrum + tooling. - RUNTIME CHEATSHEET: just verify, chatd boot, /v1/chat/providers probe, scrum_review.sh usage. - VISION: 5-loop substrate gate from project_small_model_pipeline_vision.md. The read-mem skill (in /root/.claude/skills/read-mem/) and project memory file are updated to reference this file as the primary Go anchor. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
e4ee0029c0 |
scrum_review.sh: reusable 3-lineage cross-review driver
Bash driver wrapping /v1/chat for Opus + Kimi + Qwen3-coder review runs. Used today to scrum the 4-phase wave (1,624 LoC of chatd + config-refactor + Rust cleanup) and caught 2 BLOCKs + 2 WARNs. Usage: ./scripts/scrum_review.sh <bundle.diff> <bundle_label> Output: reports/scrum/_evidence/<DATE>/verdicts/<bundle>_<reviewer>.md verbatim, per the evidence-only convention. Per-reviewer latency + token counts captured in the report header. System prompt enforces the BLOCK/WARN/INFO + WHERE/WHAT/WHY shape per feedback_cross_lineage_review.md — leads with verdict, no preamble (Kimi tends to spend tokens thinking otherwise). Reviewer fleet matches project_golang_lakehouse.md "Scrum routing": - opencode/claude-opus-4-7 - openrouter/moonshotai/kimi-k2-0905 - openrouter/qwen/qwen3-coder This is the first dogfood of chatd as the scrum vehicle — eats its own /v1/chat dispatcher. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
0efc7363c5 |
scrum 2026-04-30: 4 real fixes + 2 INFOs from cross-lineage review
3-lineage scrum (Opus 4.7 / Kimi K2.6 / Qwen3-coder) on today's wave landed 4 real findings (2 BLOCK + 2 WARN) and 2 INFO touch-ups. Verbatim verdicts + disposition table at: reports/scrum/_evidence/2026-04-30/ B-1 (BLOCK Opus + INFO Kimi convergent) — ResolveKey API: collapse from 3-arg (envVar, envFileName, envFilePath) to 2-arg (envVar, envFilePath). Pre-fix every chatd caller passed the env var name twice; if operator renamed *_key_env in lakehouse.toml while keeping the canonical KEY= line in the .env file, fallback silently missed. B-2 (WARN Opus + WARN Kimi convergent) — handleProviders probe: drop the synthesize-then-Resolve probe; look up by name directly via Registry.Available(name). Prior probe synthesized "<name>/probe" model strings and routed through Resolve, fragile to any future routing rule (e.g. cloud-suffix special case). B-3 (BLOCK Opus single — verified by trace + end-to-end probe) — OllamaCloud.Chat StripPrefix used "cloud" but registry routes "ollama_cloud/<m>". Result: upstream got the prefixed model name and 400'd. Smoke missed it because chatd_smoke runs without ollama_cloud registered. Now strips the right prefix; new TestOllamaCloud_StripsCorrectPrefix locks both prefix + suffix cases. Verified live: ollama_cloud/deepseek-v3.2 round-trips cleanly through the real ollama.com endpoint. B-4 (WARN Opus single) — Ollama finishReason: read done_reason field instead of inferring from done bool alone. Newer Ollama reports done=true with done_reason="length" on truncation; the prior code mapped that to "stop" and lost the truncation signal the playbook_lift judge needs to retry. New TestFinishReasonFromOllama_PrefersDoneReason covers the fallback ladder. INFOs: - B-5: replace hand-rolled insertion sort in Registry.Names with sort.Strings (Opus called the "avoid sort import" comment a false economy — correct). - A-1: clarify the playbook_lift.sh comment around -judge "" arg passing (Opus noted the comment said "env priority" but didn't reflect that the empty arg also passes through the Go driver's resolution chain). False positives dismissed (3, documented in disposition.md): - Kimi: TestMaybeDowngrade_WithConfigList wrong assertion (test IS correct per design — model excluded from weak list = strong = downgrade) - Qwen: nil-deref claim (defensive code already handles nil) - Opus: qwen3.5:latest doesn't exist on Ollama hub (true on the public hub but local install has it) just verify: PASS. chatd_smoke 6/6 PASS. New regression tests: 3 (B-2, B-3, B-4 each get a focused test). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
05273ac06b |
phase 4: chatd — multi-provider LLM dispatcher (ollama / cloud / openrouter / opencode / kimi)
new cmd/chatd on :3220 routes /v1/chat to the right provider based
on model-name prefix or :cloud suffix. closes the architectural gap
named in lakehouse.toml [models]: tiers map to model IDs, but until
phase 4 there was no service that could actually CALL those models
from go.
routing rules (registry.Resolve):
ollama/<m> → local Ollama (prefix stripped)
ollama_cloud/<m> → Ollama Cloud
<m>:cloud → Ollama Cloud (suffix variant — kimi-k2.6:cloud)
openrouter/<v>/<m> → OpenRouter (prefix stripped, OpenAI-compat)
opencode/<m> → OpenCode unified Zen+Go
kimi/<m> → Kimi For Coding (api.kimi.com/coding/v1)
bare names → local Ollama (default)
provider implementations:
- internal/chat/types.go Provider interface, Request/Response, errors
- internal/chat/registry.go prefix + :cloud suffix dispatch
- internal/chat/ollama.go local Ollama via /api/chat (think=false default)
- internal/chat/ollama_cloud.go Ollama Cloud via /api/generate (Bearer auth)
- internal/chat/openai_compat.go shared OpenAI Chat Completions for the
OpenRouter/OpenCode/Kimi family
- internal/chat/builder.go BuildRegistry from BuilderInput;
ResolveKey reads env then .env file fallback
config:
- ChatdConfig in internal/shared/config.go with bind, ollama_url,
per-provider key env names + .env fallback paths, timeout
- Gateway gains chatd_url + /v1/chat + /v1/chat/* routes
- lakehouse.toml [chatd] block with /etc/lakehouse/<provider>.env defaults
tests (19 in internal/chat):
- registry: prefix + :cloud + errors + telemetry + provider listing
- ollama: happy path + prefix strip + format=json + 500 mapping +
flatten_messages
- openai_compat: happy path + format=json + 429 mapping + zero-choices
think=false default in ollama + ollama_cloud — local hot path skips
reasoning, low-budget callers (the playbook_lift judge at max_tokens=10)
get direct answers instead of empty content + done_reason=length.
proven via chatd_smoke acceptance.
acceptance gate: scripts/chatd_smoke.sh — 6/6 PASS:
1. /v1/chat/providers lists exactly registered providers (1 in dev mode)
2. bare model → ollama default with content + token counts + latency
3. explicit ollama/<m> → prefix stripped at upstream
4. <m>:cloud without ollama_cloud registered → 404 (no silent fall-through)
5. unknown/<m> → falls through to default → upstream 502 (no prefix rewrite)
6. missing model field → 400
just verify: PASS (vet + 30 packages × short tests + 9 smokes).
chatd_smoke is a domain smoke (not in just verify, mirrors matrix /
observer / pathway pattern).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
848cbf5fef |
phase 3: playbook_lift harness reads judge from config
migrate the reality-test harness's judge-model default from a hardcoded "qwen3.5:latest" string to cfg.Models.LocalJudge. resolution priority: explicit -judge flag > $JUDGE_MODEL env > cfg.Models.LocalJudge from lakehouse.toml > hardcoded fallback. bumping the judge for run #N+1 now means editing one line in lakehouse.toml [models].local_judge — no Go file or shell script edits required. changes: - scripts/playbook_lift/main.go: -config flag added, judge default flips to "" so resolution chain runs. Imports internal/shared for config loader. - scripts/playbook_lift.sh: JUDGE_MODEL no longer defaulted in bash; EFFECTIVE_JUDGE resolved by mirror-of-the-Go-chain (env > config grep > qwen3.5:latest fallback). Used for the Ollama presence check + report header. Pre-flight grep avoids requiring jq just to read the toml. - reports/reality-tests/README.md: documents the 4-step priority chain. verified all 4 paths produce the expected judge: - config (no env): qwen3.5:latest (from lakehouse.toml) - env override: env wins - flag override: flag wins over env - missing config: DefaultConfig fallback still gives qwen3.5:latest just verify PASS. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
622e124b8f |
phase 2: matrix.downgrade reads WeakModels from config
migrate the strong-model auto-downgrade gate from a hardcoded weak list to cfg.Models.WeakModels. backward compatible: existing API preserved, callers that don't migrate keep using DefaultWeakModels. changes: - internal/matrix/downgrade.go: split IsWeakModel into rule-based base (`:free` suffix/infix) + literal-list lookup. New IsWeakModelInList(model, list) takes the config-supplied list. DowngradeInput grows a WeakModels field; nil falls back to DefaultWeakModels (preserves pre-phase-2 behavior). - internal/workflow/modes.go: add MatrixDowngradeWithWeakList(list) factory mirroring MatrixSearch's pattern. Plain MatrixDowngrade kept for backward compat. - cmd/matrixd/main.go: handlers struct holds weakModels populated from cfg.Models.WeakModels at startup; handleDowngrade threads it into every DowngradeInput. - cmd/observerd/main.go: registerBuiltinModes accepts weakModels and uses the factory variant. observerd reads cfg.Models.WeakModels in main(). end-to-end verified: downgrade + matrix + observer + workflow smokes all pass. Existing TestMaybeDowngrade_TruthTable + TestIsWeakModel unchanged (backward compat). Two new tests cover the config path: - TestIsWeakModelInList — covers rule + literal + empty + nil - TestMaybeDowngrade_WithConfigList — verifies cfg list overrides default Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
ec1d031996 |
phase 1: add [models] tier config — additive, no callers migrate yet
Codifies the small-model-pipeline tiering (per project_small_model_pipeline_vision.md) in lakehouse.toml [models] section. Tier names map to actual model IDs; bumping a model means editing one line, not hunting through code. Tier philosophy: - local_* : on-box Ollama. Inner-loop hot path. Repeated calls. - cloud_* : Ollama Cloud (Pro plan). Larger context, fail-up tier. - frontier_* : OpenRouter / OpenCode. Rate-limited, billed per call. weak_models is the codified "local-hot-path eligible" list — phase 2 will migrate matrix.downgrade to read it instead of hardcoding. Defaults reflect 2026-04-29 architecture: qwen3.5:latest as local (stronger than qwen2.5, same JSON-clean property), kimi-k2.6 as cloud judge (kimi-k2:1t still upstream-broken), opus-4-7 + kimi-k2-0905 as frontier review/arch via OpenRouter, opencode/claude-opus-4-7 as frontier_free leveraging the OpenCode subscription. 3 new tests in internal/shared/config_test.go: - TestDefaultConfig_ModelsTier — locks tier defaults - TestModelsConfig_IsWeak — weak-bypass list - TestLoadConfig_ModelsTOMLRoundTrip — override semantics just verify PASS (g2 had one flake on first run — Ollama transfer truncation; clean on retry, unrelated to this change). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
3dd7d9fe30 |
reality-tests: playbook-lift harness — does the 5-loop substrate beat raw cosine?
First reality test driver. Two-pass design:
- Pass 1 (cold): matrix.search use_playbook=false → small-model judge
rates top-K → record playbook entry pointing at the highest-rated
result (which may NOT be top-1 by distance — that's the discovery).
- Pass 2 (warm): same queries with use_playbook=true → measure
ranking shift. Lift = real if recorded answer becomes top-1.
Files:
- scripts/playbook_lift/main.go driver (391 LoC)
- scripts/playbook_lift.sh stack-bring-up + report gen
- tests/reality/playbook_lift_queries.txt query corpus (5 placeholders;
J writes real 20+)
- reports/reality-tests/README.md framework + interpretation
- .gitignore track reports/reality-tests/
but ignore per-run JSON evidence
This answers the gate from project_small_model_pipeline_vision.md:
"the playbook + matrix indexer must give the results we're looking
for." Without ground-truth labels, the LLM judge is the proxy — the
same small-model thesis applied to evaluation. Honest about that
limitation in the generated reports.
Driver compiles clean; full run requires Ollama + workers/candidates
ingest. Skips cleanly if Ollama absent.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
8278eb9a87 |
scrum2 cleanup: JSON-marshal in stringifyValue, drop dead detectCycle, name SourceWorkflow
5 small fixes from the §3.8 scrum2 review wave:
- workflow.stringifyValue now JSON-marshals maps/slices instead of
fmt.Sprint %v (Opus+Kimi convergent: LLM modes were getting Go's
map[k:v] syntax, which is unparseable as JSON context).
- workflow.detectCycle removed — duplicate of topoSort that discarded
the useful node ID. Validate() now calls topoSort directly and
returns its wrapped ErrCycle.
- observer.SourceWorkflow named constant — was an implicit string
cast (observer.Source("workflow")) at the cmd/observerd handler.
- Unused context imports + dead silencer comments removed across
workflow/modes.go and observerd/main.go.
- Unused store parameter dropped from registerBuiltinModes (reserved
comment removed; can be re-added when a mode actually needs it).
just verify still PASS — these are pure cleanup, no behavior change.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
c41698acae |
scrum rerun-2 — 50/60 (Δ R1 +7, Δ baseline +15) at c7e3124
Audited stash-clean |
||
|
rootandClaude Opus 4.7
|
c7e3124208 |
§3.8 second slice: real modes wired (matrix.relevance/downgrade/search,
distillation.score, drift.scorer)
Lands the workflow.Mode adapters for the §3.4 components + the
distillation scorer + drift quantifier. Workflows can now compose
real measurement capabilities; the substrate's parallel
capabilities become composable Lego bricks (per the prior commit's
closing insight).
Modes registered (in observerd's registerBuiltinModes):
Pure-function wrappers (no I/O):
- matrix.relevance → matrix.FilterChunks
- matrix.downgrade → matrix.MaybeDowngrade
- distillation.score → distillation.ScoreRecord
- drift.scorer → drift.ComputeScorerDrift
HTTP-backed:
- matrix.search → POST matrixd /matrix/search
(registered only when matrixd_url is set)
Fixture (kept from §3.8 first slice):
- fixture.echo, fixture.upper
internal/workflow/modes.go:
Each mode follows the same glue pattern: marshal generic input
through a typed struct (free schema validation + clear error
messages), call the underlying capability, return a generic
output map. Roundtrip-via-JSON gives us schema validation
without writing custom field-by-field coercion.
internal/workflow/modes_test.go (10 tests, all PASS):
- matrix.relevance filters adjacency pollution (Connector kept,
catalogd::Registry dropped — same headline as the relevance
smoke, run through the workflow mode)
- matrix.downgrade flips lakehouse→isolation on strong model;
keeps lakehouse on weak (qwen3.5:latest); errors on missing
fields
- distillation.score rates scrum_review attempt_1 as accepted;
rejects empty record
- drift.scorer reports zero drift on matched inputs; errors on
empty inputs slice
- matrix.search HTTP flow round-trips through httptest fake
matrixd; non-OK status surfaces a clear error
scripts/workflow_smoke.sh (5 assertions PASS, was 4):
New assertion #5: real-mode chain
matrix.downgrade (lakehouse + grok-4.1-fast → isolation)
→ distillation.score (scrum_review attempt_1 → accepted)
Proves §3.4 components compose through the workflow runner with
no fixture intermediation. Both nodes ran successfully, runner
recorded provenance, status=succeeded.
Mode listing assertion now expects 7 modes (5 real + 2 fixture)
instead of just the fixtures.
17-smoke regression all green. SPEC §3.8 acceptance gate G3.8.D
("Mode catalog dispatches matrix.search invocation to the matrixd
backend without going through HTTP") still pending — current path
goes through HTTP for matrix.search, which is the cleaner service-
mesh shape but slower than direct in-process. In-process dispatch
when matrixd is co-resident is a future optimization.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
e30da6e5aa |
§3.8 first slice: workflow runner skeleton + DAG executor + observerd integration
Lands the structural piece of SPEC §3.8 (Observer-KB workflow runner)
documented in
|
||
|
rootandClaude Opus 4.7
|
97dd3f826d |
SPEC §3.5/§3.6/§3.7/§3.8 — name F/B/C as port targets + add Archon-style workflow runner
Per the 2026-04-29 scope-discipline pause: the wave shipped four pieces beyond SPEC §3.4 component scope, and one architectural pattern surfaced (Archon-style multi-pass workflow runner) that's the observer's natural growth path. Document them as port targets so the next scrum review has authoritative SPEC components. §3.5 — Drift quantification (loop 5 of the PRD) Names the SCORER drift work shipped in |
||
|
rootandClaude Opus 4.7
|
bc9ab93afe |
H: observerd — autonomous-iteration witness loop (SPEC §2 port)
Port of the load-bearing pieces of mcp-server/observer.ts (Rust
system, 852 lines TS) per SPEC §2's named target. Implements PRD
loop 3 ("Observer loop — watches each run, refines configs").
Routes (all under /v1/observer/* via gateway):
GET /observer/health — liveness
GET /observer/stats — total / successes / failures /
by_source / recent_scenario_ops
(matches Rust JSON shape exactly)
POST /observer/event — record one ObservedOp; auto-defaults
timestamp + source, validates required
fields (endpoint), persists to JSONL,
appends to ring buffer
Architecture:
- internal/observer/types.go — ObservedOp model + Source taxonomy
(mcp / scenario / langfuse / overseer_correction). Mirrors the
Rust shape so JSON round-trips during cutover.
- internal/observer/store.go — Store + Persistor. Ring buffer cap
matches Rust's 2000; recent_scenarios cap matches Rust's 10.
Same persist-then-apply order as pathwayd; same corruption-
tolerant replay (skip malformed lines + warn).
- cmd/observerd — :3219 HTTP service, fronted by gateway as
/v1/observer/*.
- lakehouse.toml + DefaultConfig — [observerd] block matches the
pathwayd pattern (Bind + PersistPath; empty path = ephemeral).
Tests + smoke (all PASS):
- 7 unit tests in store_test.go: validation, default fields,
stats aggregation, recent-scenarios cap + ordering, ring-buffer
rollover at cap, JSONL round-trip persistence, corruption-
tolerant replay (1 valid + 1 corrupt + 1 valid → 2 applied)
- scripts/observer_smoke.sh: 4 assertions through gateway —
record 5 events (3 ok / 2 fail across 2 sources), stats
aggregates correctly, empty-endpoint→400, kill+restart preserves
via JSONL replay (5 ops, 3 ok, 2 err survive)
Deferred (named in package + cmd doc, not in this commit):
- POST /observer/review (cloud-LLM hand-review fall-back). The
heuristic-only path could land cheaply but the productized
cloud path (qwen3-coder fall-back) is multi-day port.
- Background loops: analyzeErrors, consolidatePlaybooks,
tailOverseerCorrections (read overseer_corrections.jsonl into
the ring buffer once per cycle).
- escalateFailureClusterToLLMTeam (failure clustering trigger
that posts to LLM Team's /api/run with code_review mode).
/relevance is NOT duplicated — already ported in
|
||
|
rootandClaude Opus 4.7
|
6392772f41 |
C: bulk playbook record — operational rating wiring
POST /v1/matrix/playbooks/bulk accepts an array of playbook entries
and records each independently — failures per-entry don't abort the
batch. Designed for two operational use cases:
1. Backfilling historical placement data into the playbook
substrate (the Rust system has 4,701 fill operations recorded
with embeddings; that data deserves to feed the Go learning
loop without a 4,701-call procedural script).
2. Batched click-tracking from a session's worth of coordinator
interactions, posted once at idle rather than per-click.
Per-entry response shape: {index, playbook_id} on success or
{index, error} on failure. Caller can inspect failures without
diffing.
Smoke (scripts/playbook_smoke.sh, new assertion #4):
Bulk POST 3 entries: 2 valid (alpha→widget-a, bravo→widget-b) +
1 invalid (empty query_text). Verifies recorded=2, failed=1,
the 2 valid ones get playbook_ids back, and the invalid one
surfaces its validation error in-line.
Single-record /matrix/playbooks/record from
|
||
|
rootandClaude Opus 4.7
|
b199093d1f |
B: matrix metadata filter — post-retrieval structured gate
Addresses the reality-test gap surfaced by the candidates and multi-corpus e2e runs ( |
||
|
rootandClaude Opus 4.7
|
be65f85f17 |
F: drift quantification — scorer drift first
PRD's 5-loop substrate names "drift" as loop 5: quantify when
historical decisions stop matching current reality. Distinct from
the rating+distillation loop because drift is MEASUREMENT, not
LEARNING. The learning loop says "this match worked, remember it";
the drift loop says "this 4-month-old playbook entry — does it
still match what the substrate would surface today?"
First-shipped drift shape: SCORER drift. When the deterministic
scorer's ScorerVersion bumps, historical ScoredRuns may no longer
match what the current scorer produces on the same EvidenceRecord.
internal/drift/drift.go:
- ScorerDriftInput — (EvidenceRecord, persisted_category) pair
- ScorerDriftEntry — one mismatch with current reasons attached
- CategoryShift — (from, to, count) cell in the shift matrix
- ScorerDriftReport — summary + sorted shift matrix + optional entries
- ComputeScorerDrift(inputs, includeEntries) — pure function;
re-runs ScoreRecord over each input and reports mismatches
Why this matters: without a drift quantifier, a scorer-rule change
silently invalidates the historical training data feeding the
learning loop. With drift quantification, a rule change surfaces
a concrete number ("847 of 4701 historical ScoredRuns now
disagree") that triggers a re-score-and-retrain cycle rather than
letting the substrate quietly rot.
Tests (6/6 PASS):
- No-drift: all 3 inputs match → 100% matched
- Shift detected: 5 inputs, 3 drift cases, drift_rate=0.6,
shift matrix shows accepted→partially_accepted x3
- Multiple shifts sorted by count desc
- includeEntries=false skips the per-mismatch list
- Empty input → all-zero report (no division-by-zero)
- ScorerVersion stamped on every report
Future drift shapes (deferred to follow-ups, named in package doc):
- PLAYBOOK drift: re-run playbook queries through current
matrix-search; recorded answer not in top-K = drift
- EMBEDDING drift: KS-test on vector distribution at T1 vs T2
- AUDIT BASELINE drift: matches Rust audit_baselines.jsonl
longitudinal signal
Pure compute. Materialization layer (read scored-runs jsonl + their
matching evidence jsonl + feed into ComputeScorerDrift) lands with
the distillation materialization commit.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
57d0df125d |
E (partial): distillation port — scorer + contamination firewall
First slice of the Rust v1.0.0 distillation substrate (e7636f2) ported to Go per ADR-001 #4 (port LOGIC, not bit-identical reproducibility). This commit lands the LOAD-BEARING pieces named in project_distillation_substrate.md memory: - The deterministic Success Scorer (8 sub-scorers + dispatch) - The contamination firewall on SFT samples (the "non-negotiable" spec property: rejected/needs_human_review NEVER ship to SFT) - All on-wire types + validators for ScoredRun, SftSample, EvidenceRecord with Provenance Files: internal/distillation/types.go — types + ScorerVersion + SftNever + ValidateScoredRun + ValidateSftSample internal/distillation/scorer.go — ScoreRecord + 8 class scorers + BuildScoredRun (deterministic) internal/distillation/scorer_test.go — ~40 test cases: - source-class dispatch (verdict / telemetry / extraction) - scrum_review (4 attempt cases) - observer_review (5 verdict cases) - audit (legacy + severity, 9 cases) - auto_apply (4 cases) - outcomes / mode_experiment / extraction - CONTAMINATION FIREWALL: ErrSftContamination sentinel fires on rejected/needs_human_review, distinct from typo errors - empty-pair guard (instruction/response trim != "") - reasons-required ScoredRun validation - deterministic sig_hash on identical input - purity check (input not mutated, repeatable output) Per the 2026-04-29 cross-lineage scrum's discipline: false-positive findings would be dismissed inline (none in this commit). Real findings would be addressed before merge — but this is greenfield port code reviewed against its Rust source line-by-line, which the test suite encodes as truth tables. Explicitly DEFERRED to follow-up commits: - Materialization layer (jsonl read/write, date-partitioned storage in data/scored-runs/YYYY/MM/DD/, evidence index) - SFT exporter (file iteration + filtering — the SCORING firewall is here; the EXPORT firewall is the next layer) - export_preference, export_rag (other export shapes) - Acceptance harness (16/16 acceptance gate that locks v1.0.0) - replay, receipts, build_evidence_index, transforms The scorer + firewall validator are pure functions — operational tooling layers on top without changing the deterministic logic the downstream learning loop depends on. The Go ScorerVersion stays at v1.0.0 to match the Rust e7636f2 baseline; bumping in the Go materialization commit is reserved for the next scoring-rule change, NOT the port itself. 15-smoke regression all green. vet clean. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
7f42089521 |
D: embed-text iteration — clean negative finding (3 variants tested)
Workers driver embed text reverted to V0 after testing 3 variants
on the "Forklift operator with OSHA-30 certification, warehouse
experience" reality-test query against 5000 workers (which contains
569 actual Forklift Operators per the
|
||
|
rootandClaude Opus 4.7
|
a730fc2016 |
scrum fixes: 4 real findings landed, 4 false positives dismissed
Cross-lineage scrum review on the 12 commits of this session
(afbb506..06e7152) via Rust gateway :3100 with Opus + Kimi +
Qwen3-coder. Results:
Real findings landed:
1. Opus BLOCK — vectord BatchAdd intra-batch duplicates panic
coder/hnsw's "node not added" length-invariant. Fixed with
last-write-wins dedup inside BatchAdd before the pre-pass.
Regression test TestBatchAdd_IntraBatchDedup added.
2. Opus + Kimi convergent WARN — strings.Contains(err.Error(),
"status 404") was brittle string-matching to detect cold-
start playbook state. Fixed: ErrCorpusNotFound sentinel
returned by searchCorpus on HTTP 404; fetchPlaybookHits
uses errors.Is.
3. Opus WARN — corpusingest.Run returned nil on total batch
failure, masking broken pipelines as "empty corpora." Fixed:
Stats.FailedBatches counter, ErrPartialFailure sentinel
returned when nonzero. New regression test
TestRun_NonzeroFailedBatchesReturnsError.
4. Opus WARN — dead var _ = io.EOF in staffing_500k/main.go
was justified by a fictional comment. Removed.
Drivers (staffing_500k, staffing_candidates, staffing_workers)
updated to handle ErrPartialFailure gracefully — print warn, keep
running queries — rather than fatal'ing on transient hiccups
while still surfacing the failure clearly in the output.
Documented (no code change):
- Opus WARN: matrixd /matrix/downgrade reads
LH_FORCE_FULL_ENRICHMENT from process env when body omits
it. Comment now explains the opinionated default and points
callers wanting deterministic behavior to pass the field
explicitly.
False positives dismissed (caught and verified, NOT acted on):
A. Kimi BLOCK on errors.Is + wrapped error in cmd/matrixd:223.
Verified false: Search wraps with %w (fmt.Errorf("%w: %v",
ErrEmbed, err)), so errors.Is matches the chain correctly.
B. Kimi INFO "BatchAdd has no unit tests." Verified false:
batch_bench_test.go has BenchmarkBatchAdd; the new dedup
test TestBatchAdd_IntraBatchDedup adds another.
C. Opus BLOCK on missing finite/zero-norm pre-validation in
cmd/vectord:280-291. Verified false: line 272 already calls
vectord.ValidateVector before BatchAdd, so finite + zero-
norm IS checked. Pre-validation is exhaustive.
D. Opus WARN on relevance.go tokenRe (Opus self-corrected
mid-finding when realizing leading char counts toward token
length).
Qwen3-coder returned NO FINDINGS — known issue with very long
diffs through the OpenRouter free tier; lineage rotation worked
as designed (Opus + Kimi between them caught everything Qwen
would have).
15-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix, relevance, downgrade, playbook).
Unit tests all green (corpusingest +1, vectord +1).
Per feedback_cross_lineage_review.md: convergent finding #2 (404
detection) is the highest-signal one — both Opus and Kimi
flagged it independently. The other Opus findings stand on
single-reviewer signal but each one verified against the actual
code.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
06e71520c4 |
matrix: playbook memory + boost — SPEC §3.4 component 5 of 5 (LEARNING LOOP)
Closes SPEC §3.4. The matrix indexer is now a learning meta-index per
feedback_meta_index_vision.md — every successful (query → answer)
pair recorded via /matrix/playbooks/record boosts that answer for
future similar queries.
This is the architectural piece that lifts vectord from "static
hybrid search" to the meta-index J originally framed in Phase 19 of
the Rust system.
What's new:
- internal/matrix/playbook.go — PlaybookEntry, PlaybookHit,
ApplyPlaybookBoost. Pure-function boost math:
distance' = distance * (1 - 0.5 * score)
Score 0 = no boost (factor 1.0); score 1 = halve distance
(factor 0.5). Capped at 0.5 deliberately so a single high-
confidence playbook can't dominate the base ranking forever
(runaway-feedback-loop guard).
- Retriever.Record(entry, corpus) — embeds query_text, ensures
playbook corpus exists (idempotent), upserts via deterministic
sha256-derived ID (last score wins on re-record of same triple).
- Retriever.Search extended with UsePlaybook + PlaybookCorpus +
PlaybookTopK + PlaybookMaxDistance. Reuses the query vector —
no extra embed call. Missing-corpus 404 = no-op (cold-start
state before any Record call), not an error.
- POST /v1/matrix/playbooks/record (matrixd) — caller submits
{query_text, answer_id, answer_corpus, score, tags?}; gets
{playbook_id} back.
Storage: a vectord index named "playbook_memory" (configurable per
request) with embed(query_text) as the vector and the
PlaybookEntry JSON as metadata. Just another corpus — observable
from /vectors/index, persistable through G1P, etc.
Match key for boost: (AnswerID, AnswerCorpus). Cross-corpus ID
collisions don't false-match — verified by
TestApplyPlaybookBoost_CorpusAttributionRespected.
End-to-end smoke (scripts/playbook_smoke.sh, all assertions PASS):
- Baseline search: widget-c at distance 0.6566 (rank 3)
- Record playbook: query → widget-c, score=1.0
- Re-search with use_playbook=true:
widget-c distance: 0.3283 (rank 2)
ratio: 0.5 EXACTLY (matches boost math precisely)
playbook_boosted: 1
- widget-c jumped from #3 to #2 — learning loop visible
Tests:
- 8 unit tests in internal/matrix/playbook_test.go covering
Validate, BoostFactor (5 cases), the no-boost identity, the
boost-moves-result-up scenario, highest-score wins on duplicate
matches, cross-corpus attribution, JSON round-trip, and
rejection of empty metadata
- scripts/playbook_smoke.sh integration test (3 assertions PASS)
15-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix, relevance, downgrade, playbook).
SPEC §3.4 NOW COMPLETE: 5 of 5 components shipped. The matrix
indexer's port is done as a substrate; remaining work is operational
(rating signal sources, telemetry, eventual structured filtering for
staffing data — none in §3.4).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
31b408882b |
multi_corpus_e2e: WORKERS_LIMIT env knob — and the embed-text-not-sample-size finding
Adds WORKERS_LIMIT env override (default 5000) so the e2e can be
re-run at different sample sizes. Tiny change; the interesting part
is the FINDING that motivated the run.
Investigation: a97881d's reality test put zero Forklift Operators in
the top-6 for "Forklift operator with OSHA-30 certification,
warehouse experience" — instead returned Production Worker / Machine
Operator / Assembler.
Hypothesis tested: maybe the 5000-row sample didn't contain
forklift operators in retrievable density.
Result: hypothesis falsified. Direct probe of workers_500k.parquet:
All 500K rows → 55,349 Forklift Operators (11.07%)
→ 150,328 with "forklift" in certs
→ 74,852 with OSHA-30 specifically
First 5K rows → 569 Forklift Operators (11.38%)
→ distribution matches global, no ordering bias
So 569 forklift operators were IN the corpus the matrix indexer
searched and STILL didn't surface in top-6. That means the bottleneck
isn't sample size — it's nomic-embed-text + our embed-text template
ranking "Production Worker" / "Machine Operator" / "Assembler" as
semantically nearer to the query than literal "Forklift Operator".
The reality test exposed this faithfully. Three real follow-ups, none
in scope of this commit:
1. Embed text design — front-loading role + certs (currently
"Worker role: <role>" then skills then certs) might dominate
retrieval better. Worth A/B-testing.
2. Hybrid SQL+semantic — pre-filter by role/certs via queryd
before semantic ranking. Not in SPEC §3.4 today; would address
the "available" / "Chicago" gap from the candidates reality
test (
|
||
|
rootandClaude Opus 4.7
|
a97881d80c |
workers corpus + multi-corpus reality test — matrix indexer end-to-end
Lands the second real-data corpus (workers_500k) and the first
multi-corpus reality test through /v1/matrix/search composing both
corpora live.
What's new:
- scripts/staffing_workers/main.go — parquet driver over
workers_500k.parquet, multi-chunk arrow handling (workers
parquet has multiple row groups vs candidates' one). Embed text:
role + skills + certifications + city + state + archetype +
resume_text. IDs prefixed "w-".
- scripts/multi_corpus_e2e.sh — first end-to-end test composing
both corpora through the matrix indexer.
Real-data multi-corpus result (this commit):
Query: "Forklift operator with OSHA-30 certification, warehouse
experience"
Corpora: workers (5000 rows) + candidates (1000 rows)
Merged top-8: workers=6, candidates=2
Top hits:
w d=0.327 w-4573 Production Worker
w d=0.353 w-1726 Machine Operator
w d=0.362 w-3806 Production Worker
w d=0.366 w-1000 Machine Operator
w d=0.374 w-1436 Assembler
w d=0.395 w-162 Machine Operator
c d=0.440 c-CAND-00727 C#,.NET,Azure
c d=0.446 c-CAND-00031 React,TypeScript,Node
The matrix indexer correctly chose the right domain — manufacturing/
warehouse roles in workers (correct semantic match for the staffing
query) rank ABOVE software-engineer candidates from the candidates
corpus. 0.11 gap between the worst worker (0.395) and the best
candidate (0.440) — clean distance separation.
Compared to the candidates-only e2e run from
|
||
|
rootandClaude Opus 4.7
|
3968ec8a7b |
matrix: strong-model downgrade gate — SPEC §3.4 component 4 of 5
Pure-Go port of mode.rs::execute's pass5 downgrade gate (Rust
2026-04-26). Adds POST /v1/matrix/downgrade endpoint via matrixd.
The gate captures the pass5 finding: composing matrix corpora into
codereview_lakehouse on a strong model LOST 5/5 head-to-head reps
against matrix-free codereview_isolation on grok-4.1-fast (p=0.031).
Strong models have enough native capacity that bug fingerprints +
adversarial framing + file content carry them; matrix chunks
displace depth-of-analysis.
Logic (matches Rust mode.rs:614-632):
if mode == codereview_lakehouse
&& !forced_mode
&& !LH_FORCE_FULL_ENRICHMENT
&& !is_weak_model(model)
→ flip to codereview_isolation, record downgraded_from
is_weak_model captures the empirical weak-list:
- `:free` suffix or `:free/` infix (OpenRouter free tier)
- qwen3.5:latest, qwen3:latest (local last-resort rungs)
- everything else → strong by default
Tests:
- 3 unit tests in internal/matrix/downgrade_test.go: IsWeakModel
coverage, MaybeDowngrade truth table (5 rows), forced-mode
precedence (forced beats every other bypass)
- scripts/downgrade_smoke.sh: 6 assertions through gateway covering
all 5 truth-table rows + empty-mode 400
14-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix, relevance, downgrade).
SPEC §3.4 progress: 4 of 5 components shipped (corpus builders,
multi-corpus retrieve+merge, relevance filter, downgrade gate).
Last component is learning-loop integration.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
9588bd82ae |
matrix: relevance filter — SPEC §3.4 component 3 of 5
Faithful port of mcp-server/relevance.ts (Rust observer's adjacency-
pollution filter). Same 5-signal scoring, same default threshold 0.3.
Adds POST /v1/matrix/relevance endpoint via matrixd.
Scoring signals (additive, can sign-flip):
path_match +1.0 chunk source/doc_id encodes focus.path
filename_match +0.6 chunk text mentions focus's filename
defined_match +0.6 chunk text mentions focus.defined_symbols
token_overlap +0.4 jaccard of non-stopword tokens
prefix_match +0.3 chunk source shares first-2-segment prefix
import_penalty -0.5 mentions ONLY imported symbols, no defined ones
What this does and doesn't do:
- DOES filter code-aware corpora (eventually lakehouse_arch_v1,
lakehouse_symbols_v1, scrum_findings_v1) — drops chunks about
code the focus file IMPORTS rather than DEFINES, the
"adjacency pollution" pattern that makes a reviewer LLM
hallucinate imported-crate internals as belonging to the focus
- DOES NOT meaningfully filter staffing data — the candidates
reality test 2026-04-29 had "exact skill match buried at #3"
which is a different problem (semantic-only ranking dominated
by secondary text). Staffing needs structured filtering
(status gates, location gates) that lives outside this
package — future work, not in SPEC §3.4 yet
Headline smoke assertion: focus = crates/queryd/src/db.go which
defines Connector and imports catalogd::Registry. The filter
scores:
Connector chunk: +0.68 (defined_match fires, kept)
Registry chunk: -0.46 (import_only penalty fires, dropped)
unrelated junk: 0.00 (no signals, dropped)
That's a 1.14-point gap between what we ARE and what we IMPORT —
the entire purpose of the filter.
Tests:
- 9 unit tests in internal/matrix/relevance_test.go covering
Tokenize, Jaccard, ExtractDefinedSymbols (Rust + TS),
ExtractImportedSymbols, FilePrefix, ScoreRelevance per-signal,
FilterChunks threshold splitting, and the headline
AdjacencyPollutionScenario
- scripts/relevance_smoke.sh integration smoke (3 assertions PASS):
adjacency-pollution scenario, empty-chunks 400, threshold honored
13-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix, relevance).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
0d1553ca88 |
candidates corpus: first deep-field reality test on real staffing data
Lands the second staffing corpus and the first end-to-end reality test
through the full Go pipeline: parquet → corpusingest → embedd →
vectord → matrixd → gateway.
What's new:
- scripts/staffing_candidates/main.go — parquet Source over
candidates.parquet (1000 rows, 11 cols), single-chunk arrow-go
pqarrow read. Embed text: "Candidate skills: <s>. Based in
<city>, <state>. <years> years experience. Status: <status>.
<first> <last>." IDs prefixed "c-" so multi-corpus merges
against workers ("w-") stay unambiguous.
- scripts/candidates_e2e.sh — first integration smoke that runs
the full stack (storaged + embedd + vectord + matrixd + gateway),
ingests via corpusingest, runs a real query through
/v1/matrix/search, prints results. Ephemeral mode (vectord
persistence disabled via custom toml) so re-runs don't pollute
MinIO _vectors/ and break g1p_smoke's "only-one-persisted-index"
assertion.
Real bug caught + fixed in corpusingest:
When LogProgress > 0, the progress goroutine's only exit was
ctx.Done(). With context.Background() in the production driver,
Run hung forever after the pipeline finished. Added a stopProgress
channel that close()s after wg.Wait(). Regression test
TestRun_ProgressLoggerExits bounds Run's wall to 2s with
LogProgress=50ms.
This is the bug the unit tests didn't catch because every prior test
set LogProgress: 0. Reality test surfaced it on first real-data
run — exactly the hyperfocus-and-find-architectural-weakness
property J framed as the reason for the Go pass.
End-to-end output (1000 candidates, query "Python AWS Docker
engineer in Chicago available now"):
populate: scanned=1000 embedded=1000 added=1000 wall=3.5s
matrix returned 5 hits in 26ms
The result quality is the interesting signal: top-5 had ZERO
Chicago candidates, ZERO active-status candidates, and the exact-
skill-match (Python,AWS,Docker) ranked #3 not #1. Pipeline works;
retrieval quality has real architectural limits (no structured
filtering, no relevance gate, semantic-only ranking dominated by
secondary signals like "1 year experience" and "engineer"). This
motivates SPEC §3.4 components 3 (relevance filter) and
eventually structured filtering — exactly the kind of finding the
deep field reality tests are supposed to surface before Enterprise
cutover.
12-smoke regression sweep all green. 9 corpusingest unit tests
including the new regression. vet clean.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
166470f532 |
corpusingest: extract reusable text→vector ingest pipeline
Generalizes the staffing_500k driver's embed-and-push loop into
internal/corpusingest. Per docs/SPEC.md §3.4 component 1 (corpus
builders): adding a new staffing/code/playbook corpus is now one
Source impl + one main.go calling Run, not 200 lines of pipeline
copy-paste.
API:
type Source interface { Next() (Row, error) }
func Run(ctx, Config, Source) (Stats, error)
Library owns:
- Index lifecycle (create, optional drop-existing, idempotent
reuse on 409)
- Parallel embed dispatcher (configurable workers + batch size)
- Vectord push batching
- Progress logging + Stats reporting
- Partial-failure semantics (log + continue per-batch errors;
operator decides on re-run via Stats.Embedded vs Scanned delta)
Per-corpus driver owns: source parsing + column→Row mapping +
post-ingest validation queries.
Refactor scripts/staffing_500k/main.go to use it. Driver is now
~190 lines (was 339), with the embed/add plumbing replaced by one
Run call. -drop flag added so callers can opt out of the destructive
DELETE-first behavior (default still true to keep the 500K test
clean-recall semantics).
Unit tests (internal/corpusingest/ingest_test.go, 8/8 PASS):
- Pipeline shape: 50 rows / 16 batch → 4 embed + 4 add calls,
every ID added exactly once, vectors at correct dimension
- DropExisting fires DELETE
- 409 on create → reuse existing index
- Limit stops early
- Empty Text rows skipped (counted as scanned, not added)
- Required IndexName + Dimension validation
- Context cancel stops mid-pipeline
Real bug caught and fixed by the test suite: if embedd ever returns
fewer vectors than texts in the request (degraded backend), the
addBatch loop would panic with index-out-of-range. Worker now
length-checks the response and logs+skips on mismatch.
12-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix). vet clean.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
c1d96b7b60 |
matrixd: multi-corpus retrieve+merge — SPEC §3.4 component 2 of 5
Lands the matrix indexer's first piece per docs/SPEC.md §3.4:
multi-corpus retrieve+merge with corpus attribution per result.
Future components (relevance filter, downgrade gate, learning-loop
integration) layer on top of this surface.
Architecture:
- internal/matrix/retrieve.go — Retriever takes (query, corpora,
k, per_corpus_k), parallel-fans across vectord indexes, merges
by distance ascending, preserves corpus origin per hit
- cmd/matrixd — HTTP service on :3217, fronts /v1/matrix/*
- gateway proxy + [matrixd] config + lakehouse.toml entry
- Either query_text (matrix calls embedd) or query_vector
(caller pre-embedded) — vector takes precedence if both set
Error policy: fail-loud on any corpus error. Silent partial returns
would lie about coverage, defeating the matrix's whole purpose.
Bubbles vectord errors as 502 (upstream), validation as 400.
Smoke (scripts/matrix_smoke.sh, 6 assertions PASS first try):
- /matrix/corpora lists indexes
- Multi-corpus search returns hits from BOTH corpora
- Top hit is the globally-closest across all corpora
(b-near beats a-near at distance 0.05 vs 0.1 — proves merge)
- Metadata round-trips through the merge
- Distances ascending in result list
- Negative paths: empty corpora → 400, missing corpus → 502,
no query → 400
12-smoke regression sweep all green (D1-D6, G1, G1P, G2,
storaged_cap, pathway, matrix).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
a7620c8b6f |
PRD: name the product vision — small-model pipeline + 5-loop substrate
Adds a "Product vision" section before the Direction-pivot section. Captures the framing J flagged 2026-04-29: the Go refactor is not the goal. The goal is a small-model-driven autonomous pipeline that gets better with each run, with frontier models in audit/oversight, not the hot path. Five loops named explicitly: 1. Knowledge pathway (pathway memory + matrix indexer) 2. Execution (small models on focused context) 3. Observer (refines configs that got the model to a good pathway) 4. Rating + distillation (outcomes fold back into the playbook) 5. Drift (measure when the playbook stops matching reality) Triage / human-in-loop named as the system's job, not an escape hatch. The gate: "playbook + matrix indexer must give the results we're looking for" — single load-bearing acceptance criterion. Why Go after Rust: second-language pass surfaces architectural weaknesses Rust hid; the pipeline must work AS A PIPELINE, not as crates that interact. Maps existing Rust components (✓ pathway, ✓ matrix, ✓ observer, ✓ distillation, ✓ auditor; partial: drift, rating gate, triage). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
71b35fb85e |
SPEC §1 + §3.4: name matrix indexer as a port target
Adds matrix indexer as its own row in the §1 component table and a new §3.4 with port plan. Distinct from vectord (substrate); lives at internal/matrix/ + gateway /v1/matrix/*. Five components in dependency order: corpus builders → multi-corpus retrieve+merge → relevance filter → strong-model downgrade gate → learning-loop integration. Locks in the framing J flagged 2026-04-29: in Rust the matrix indexer was emergent across mode.rs + build_*_corpus.ts + observer /relevance, and earlier port-planning reduced it to "we have vectord." The SPEC now names it explicitly so the port preserves the multi-corpus retrieval shape AND the learning loop, not just the HNSW substrate. Sharding-by-id was investigated as a throughput fix and rejected — corpus-as-shard at the matrix layer is the existing retrieval shape and parallelizes Adds for free. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
f1c188323c |
vectord: BatchAdd — single-lock variadic batch (Option A)
Replaces the per-item Add loop in the HTTP handler with one call to Index.BatchAdd, which acquires the write-lock once and pushes the whole batch through coder/hnsw's variadic Graph.Add. Pre-validation stays in the handler so per-item error messages keep their item-index precision. Microbench (internal/vectord/batch_bench_test.go) at d=768 cosine: N=16 SingleAdd 283µs/op → BatchAdd 170µs/op 1.66× N=128 SingleAdd 7.9ms/op → BatchAdd 7.5ms/op 1.05× N=1024 SingleAdd 87.5ms/op → BatchAdd 83.4ms/op 1.05× Win is biggest at staffing-driver batch sizes (N=16) where per-call lock + validation overhead is a meaningful fraction. At larger N the inner HNSW neighborhood search per insert dominates, which is the load-bearing finding for Option B (sharded indexes): the throughput ceiling lives inside the library, not at the lock, so sharding to N parallel Graphs is the only path to true concurrent-Add throughput. g1, g1p, g2 smokes all PASS post-change. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
afbb506dbc |
pathwayd: HTTP service over internal/pathway · 11/11 smoke gate
Network-callable Mem0-style trace memory at :3217, fronted by gateway
/v1/pathway/*. Closes the ADR-004 wire-up: store substrate landed in
|
||
|
rootandClaude Opus 4.7
|
2a6234ff82 |
ADR-004 + internal/pathway: Mem0 versioned trace substrate
Closes Sprint 2 design-bar work (audit reports/scrum/sprint-backlog.md):
S2.1 — ADR-004 documents the pathway-memory data model
S2.2 — pathway port lands with deterministic fixture corpus
and full test coverage on day one
S2.3 — retired traces are excluded from retrieval (test
passes; would fail without the filter)
Mem0-style operations: Add / AddIdempotent / Update / Revise /
Retire / Get / History / Search. Each operation is a method on
Store; persistence is JSONL append-only with corruption recovery
on Replay.
internal/pathway/types.go Trace + event + SearchFilter + sentinel errors
internal/pathway/store.go in-memory state + RWMutex + ops
internal/pathway/persistor.go JSONL append-only log with replay
internal/pathway/store_test.go 20 test funcs covering all 7
Sprint 2 claim rows + concurrency
internal/pathway/persistor_test.go 6 test funcs covering missing-
file, corruption recovery, long-line
handling, parent-dir auto-create,
apply-error skip behavior
Sprint 2 claim coverage row-by-row:
ADD TestAdd_AssignsUIDAndTimestamps + TestAdd_RejectsInvalidJSON
UPDATE TestUpdate_ReplacesContentSameUID + Update_MissingUID_Errors
REVISE TestRevise_LinksToPredecessorViaHistory +
TestRevise_PredecessorMissing_Errors +
TestRevise_ChainOfThree_BackwardWalk
RETIRE TestRetire_ExcludedFromSearch +
TestRetire_StillAccessibleViaGet +
TestRetire_StillAccessibleViaHistory
HISTORY/cycle TestHistory_CycleDetected (injected via internal map),
TestHistory_PredecessorMissing_TruncatesChain,
TestHistory_UnknownUID_ErrorsClean
REPLAY/dup TestAddIdempotent_IncrementsReplayCount (locks the
"replay preserves original content" rule per ADR-004)
CORRUPTION TestPersistor_CorruptedLines_Skipped +
TestPersistor_ApplyError_Skipped
ROUND-TRIP TestPersistor_RoundTrip locks the full Save → fresh
Store → Load → Stats-match contract
Two real bugs caught during testing:
- Add returned the same *Trace stored in the map, so callers
holding a reference saw later mutations. Fixed: clone before
return (matches Get's contract). Same fix in AddIdempotent
+ Revise.
- Test typo: {"v":different} isn't valid JSON; AddIdempotent's
json.Valid rejected it as ErrInvalidContent. Test fixed to
use {"v":"different"}; the validation behavior is correct.
Skipped this commit (next):
- cmd/pathwayd HTTP binary
- gateway routing for /v1/pathway/*
- end-to-end smoke
These add the wire surface; the substrate ships first so the
wire layer can be a pure proxy in the next commit.
Verified:
go test -count=1 ./internal/pathway/ — 26 tests green
just verify — vet + test + 9 smokes 34s
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
ad1670d36a |
storaged cap smoke — verifies ADR-002 at 300 MiB
Closes the "needs heavy integration smoke" follow-up from the
ADR-002 commit (
|
||
|
rootandClaude Opus 4.7
|
fa56134b90 |
ADR-003 wiring: Bearer token + IP allowlist middleware
Implements the auth posture from ADR-003 (commit |
||
|
rootandClaude Opus 4.7
|
8f4c16fab1 |
mcpd: Go MCP SDK port — replaces Bun mcp-server tool surface
New cmd/mcpd binary using github.com/modelcontextprotocol/go-sdk
v1.5.0 over stdio transport. Exposes Lakehouse capabilities as MCP
tools: list_datasets, get_manifest, query_sql, embed_text,
search_vectors. Each tool proxies to the gateway via HTTP.
Replaces the MCP-tool subset of the Rust system's Bun mcp-server.ts
(the audit's "split this 2520-line empire" finding from R-005). HTTP
demo routes (the staffing co-pilot UI at /api/intelligence/*,
/headshots/*, etc.) stay Bun until G5 cutover — those are demo-
specific and depend on matrix-indexer signals not yet ported.
Architecture:
cmd/mcpd/main.go (235 LoC)
main() reads --gateway flag, builds server via buildServer(),
runs on StdioTransport. Each tool's args is a typed struct with
jsonschema tags (the SDK's canonical pattern); reflection
generates the JSON Schema automatically.
gatewayClient: thin HTTP wrapper over the configured gateway URL.
30s per-request timeout. 16 MiB tool-response cap. Non-2xx
surfaces as IsError CallToolResult (NOT as transport error) so
the LLM caller sees the error text and can decide how to react.
proxy() handles GET + POST + JSON body uniformly. errorResult()
+ jsonResult() helpers normalize CallToolResult shape.
cmd/mcpd/main_test.go (13 test funcs)
Tests the full MCP wire end-to-end without a subprocess: spin
up a fake gateway via httptest, build the MCP server pointed at
it, connect a client via in-memory transports (NewInMemoryTransports),
call each tool. Each tool gets:
- happy path (gateway returns 200 → tool returns content)
- input validation (missing required fields → IsError)
- upstream error (gateway 4xx → tool returns IsError)
Plus TestListTools verifies all 5 tools register; TestGatewayUnreachable
verifies network-level failures surface as IsError, not panics.
Setup for Claude Desktop / Code documented in README:
{
"mcpServers": {
"lakehouse": {
"command": "/path/to/bin/mcpd",
"args": ["--gateway", "http://127.0.0.1:3110"]
}
}
}
Verified:
go test -count=1 ./cmd/mcpd/ — 13/13 green
just verify — vet + test + 9 smokes 35s
Out of scope for this commit:
- Resources (mcp.AddResource): not needed yet; tools cover the
interactive surface. Add when an LLM-side use case shows up.
- Prompts (mcp.AddPrompt): same.
- Streamable transports (HTTP, SSE): stdio is the universal one;
streamable can be added with srv.Run(ctx, &mcp.StreamableHTTPHandler{})
swap if a daemon-mode deploy makes sense.
- mcpd inside the daemon-supervised stack: it's stdio-only and
spawned by the MCP client, not run as a service. Adding a
daemon-mode (HTTP transport on a port) is a follow-up if MCP
consumers want long-lived sessions.
This is a tool-surface only port. The Bun mcp-server.ts also serves
HTTP demo routes (/api/catalog/datasets, /intelligence/*, /headshots/*)
that depend on the matrix-indexer signals from the Rust system; those
stay Bun until G5 cutover when the staffing co-pilot service ports
to Go.
Direct deps added:
github.com/modelcontextprotocol/go-sdk v1.5.0
Transitive (resolved by go mod tidy):
github.com/google/jsonschema-go v0.4.2
github.com/yosida95/uritemplate/v3 v3.0.2
golang.org/x/oauth2 v0.35.0
github.com/segmentio/encoding v0.5.4
github.com/golang-jwt/jwt/v5 v5.3.1
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
56844c3f31 |
embed cache — LRU at /v1/embed for repeat-query elimination
Adds CachedProvider wrapping the embedding Provider with a thread-safe
LRU keyed on (effective_model, sha256(text)) → []float32. Repeat
queries return the stored vector without round-tripping to Ollama.
Why this matters: the staffing 500K test (memory project_golang_lakehouse)
documented that the staffing co-pilot replays many of the same query
texts ("forklift driver IL", "welder Chicago", "warehouse safety", etc).
Each repeat paid the ~50ms Ollama round-trip. Cached repeats now serve
in <1µs (LRU lookup + sha256 of input).
Memory budget: ~3 KiB per entry at d=768. Default 10K entries ≈ 30 MiB.
Configurable via [embedd].cache_size; 0 disables (pass-through mode).
Per-text caching, not per-batch — a batch with mixed hits/misses only
fetches the misses upstream, then merges the result preserving caller
input order. Three-text batch with one miss = one upstream call for
that one text instead of three.
Implementation:
internal/embed/cached.go (NEW, 150 LoC)
CachedProvider implements Provider; uses hashicorp/golang-lru/v2.
Key shape: "<model>:<sha256-hex>". Empty model resolves to
defaultModel (request-derived) for the key — NOT res.Model
(upstream-derived), so future requests with same input shape
hit the same key. Caught by TestCachedProvider_EmptyModelResolvesToDefault.
Atomic hit/miss counters + Stats() + HitRate() + Len().
internal/embed/cached_test.go (NEW, 12 test funcs)
Pass-through-when-zero, hit-on-repeat, mixed-batch only fetches
misses, model-key isolation, empty-model resolves to default,
LRU eviction at cap, error propagation, all-hits synthesized
without upstream call, hit-rate accumulation, empty-texts
rejected, concurrent-safe (50 goroutines × 100 calls), key
stability + distinctness.
internal/shared/config.go
EmbeddConfig.CacheSize (toml: cache_size). Default 10000.
cmd/embedd/main.go
Wraps Ollama Provider with CachedProvider on startup. Adds
/embed/stats endpoint exposing hits / misses / hit_rate / size.
Operators check the rate to confirm the cache is working
(high rate = good) or sized wrong (low rate + many misses on a
workload that should have repeats).
cmd/embedd/main_test.go
Stats endpoint tests — disabled mode shape, enabled mode tracks
hits + misses across repeat calls.
One real bug caught by my own test:
Initial implementation cached under res.Model (upstream-resolved)
rather than effectiveModel (request-resolved). A request with
model="" caching under "test-model" (Ollama's default), then a
request with model="the-default" (our config default) missing
the cache. Fix: always use the request-derived effectiveModel
for keys; that's the predictable side. Locked by
TestCachedProvider_EmptyModelResolvesToDefault.
Verified:
go test -count=1 ./internal/embed/ — all 12 cached tests + 6 ollama tests green
go test -count=1 ./cmd/embedd/ — stats endpoint tests green
just verify — vet + test + 9 smokes 33s
Production benefit:
~50ms Ollama round-trip → <1µs cache lookup for cached entries.
At 10K-entry default + ~30% repeat rate (typical staffing co-pilot
workload), saves several seconds per staffer-query session.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
fb08232f58 |
Batch 4: embed fixture-mode — partial R-006 closure
Adds cmd/fake_ollama, a minimal Ollama-API-compatible fake that
implements just enough surface for embedd to drive end-to-end
without a real Ollama install:
GET /api/tags — fixed model list including nomic-embed-text
POST /api/embeddings — deterministic dim-D vector from sha256(prompt)
GET /health — for the smoke's poll_health helper
Same prompt → bit-identical vector across runs, machines, and CI
nodes. Vectors are NOT semantically meaningful; the fake validates
the embed CONTRACT (dimension echo, response shape, status codes,
deterministic round-trip), not real semantic ranking. Real ranking
still requires real Ollama and lives in scripts/g2_smoke.sh + the
integration tier of the proof harness.
scripts/g2_smoke_fixtures.sh — full chain smoke against the fake:
- Build fake_ollama + embedd + vectord + gateway
- Start fake on :11435 (distinct from real Ollama at :11434)
- Generate temp lakehouse.toml with provider_url override
- Boot embedd/vectord/gateway with --config <override>
- 4 assertions: dim=768, deterministic same-text, different-text
divergence, bad-model → 4xx/5xx (fake 404 → embedd 502)
- Trap-cleanup tears down all 4 binaries + tmp config
Wired into the task runner:
just smoke-g2-fixtures
Closes R-006 partially:
- Embed half: ✓ — CI / fresh-clone reviewers without Ollama can
now run the embed contract smoke
- Storage half: deferred — mocking S3 protocol is non-trivial
(multipart, signed URLs, etc.) and MinIO itself is lightweight
enough to install via Docker in any CI environment. Documented
as Sprint 0 follow-up if a CI system without Docker shows up.
What this DOESN'T cover:
- Real semantic similarity (use scripts/g2_smoke.sh + real Ollama)
- Real Ollama API quirks (timeouts, version-specific shapes,
/api/embed batch endpoint that newer versions support)
Verified:
bash scripts/g2_smoke_fixtures.sh — 4/4 assertions PASS, ~3s wall
just verify — vet + test + 9 smokes still green
Doesn't replace the existing g2_smoke.sh (which still requires real
Ollama and exercises the actual embed semantics). Adds an alternate
mode for portability.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
0f79bce948 |
Batch 3: cmd/<bin>/main_test.go × 6 — closes R-005
Adds main_test.go for each of the 6 cmd binaries that lacked them
(storaged already had main_test.go; that's where the pattern came
from). Each test file focuses on the cmd-specific surface — route
mounts, body caps, decode/validation paths — without re-testing
internal package logic that's covered elsewhere.
cmd/catalogd/main_test.go — 6 funcs
TestRoutesMounted: chi.Walk asserts /catalog/{register,manifest/*,list}
TestHandleRegister_BodyTooLarge: 5 MiB body → 4xx
TestHandleRegister_MalformedJSON: 400
TestHandleRegister_EmptyName_400: ErrEmptyName surfaces as 400
TestHandleGetManifest_404 + TestHandleList_EmptyShape
cmd/embedd/main_test.go — 8 funcs
stubProvider implements embed.Provider deterministically
TestRoutesMounted, MalformedJSON_400, EmptyTextRejected_400 (per
scrum O-W3), UpstreamError_502 (provider error → 502, not 500),
HappyPath_ProviderEcho, BodyTooLarge (4xx range), TestItoa
(covers the no-strconv helper)
cmd/gateway/main_test.go — 4 funcs
TestMustParseUpstream_HappyPaths: 3 valid URLs
TestMustParseUpstream_FailureExits: re-execs the test binary in a
subprocess with env flag (standard pattern for testing os.Exit
callers); subprocess invokes mustParseUpstream("127.0.0.1:3211")
[missing scheme]; expects exit non-zero. Same pattern for garbage.
TestUpstreamConfigKeys_DocumentedShape: locks the 6 _url keys
cmd/ingestd/main_test.go — 7 funcs
Stubs both storaged and catalogd via httptest.Server so the cmd
layer can be exercised without bringing the full chain up.
TestHandleIngest_MissingNameQueryParam: 400 with "name" in body
TestHandleIngest_MalformedMultipart: 400
TestHandleIngest_MissingFormFile: 400 (valid multipart, wrong field)
TestHandleIngest_BodyTooLarge: 4xx
TestEscapeKeyPath: 6-case URL-escape table (apostrophe, space, etc.)
TestParquetKeyPath_Format: locks the datasets/<n>/<fp>.parquet shape
per scrum C-DRIFT (any rename breaks idempotent re-ingest)
cmd/queryd/main_test.go — 6 funcs
Tests pre-DB paths (decode, body cap, empty SQL); db.QueryContext
itself needs DuckDB so it's covered by GOLAKE-040 in the proof
harness, not unit tests. handlers.db = nil here is intentional.
TestHandleSQL_EmptySQL_400: 3 cases (empty, whitespace, mixed-WS)
TestMaxSQLBodyBytes_Reasonable: locks the 64 KiB constant in a
sane range so a refactor can't blow it open
TestPrimaryBucket_Constant: locks "primary" — secrets lookup uses
this; rename = silent secret-resolution failure at boot
cmd/vectord/main_test.go — 14 funcs
All 6 routes verified mounted. handlers.persist = nil = pure
in-memory mode; persistence is GOLAKE-070 in the proof harness.
Coverage of every error branch in handleCreate/Add/Search/Delete:
missing index → 404, dim mismatch → 400, empty items → 400,
empty id → 400, malformed JSON → 400, body too large → 4xx,
happy create → 201, happy list → 200.
One real finding caught during writing:
Body-cap rejection is sometimes 413 (typed MaxBytesError survives
unwrap) and sometimes 400 (decoder wraps it as a generic decode
error). Both are valid client-error contracts; the contract isn't
"exactly 413" but "fails loud as 4xx, never silent 200 or 5xx."
Tests assert 4xx range. The proof harness's
proof_assert_status_4xx already had this shape — just bringing
the unit tests in line with it.
Verified:
go test -count=1 -short ./cmd/... — all 7 packages green
just verify — vet + test + 9 smokes 35s
Closes audit risk R-005 (6/7 cmd/main.go untested). Combined with
the proof harness's wiring coverage, every cmd-level handler now
has both unit-test and integration-test coverage of the wiring
layer. R-005 → CLOSED.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
1ec85b0a16 |
Batch 2: perf baseline — multi-sample + warmup + MAD threshold
Replaces single-shot baselines (40% noise floor flagged in Phase E)
with noise-aware regression detection.
What changed:
ingest n=3 runs (was 1) with 3-pass warmup
vector_add n=3 runs (was 1) with 3-pass warmup
query n=20 samples (unchanged) with 50-pass warmup
search n=20 samples (unchanged) with 50-pass warmup
RSS n=1 (unchanged — steady-state in G0)
Each metric stored as {value: median, mad: median absolute
deviation} in baseline.json (schema: v2-multisample-mad).
New regression detection:
threshold = max(3 * baseline.mad, value * 0.75)
REGRESSION iff |actual - baseline.value| > threshold AND direction
signals worse (lower throughput / higher latency).
Why these specific numbers:
3*MAD = standard "outside the spread" bound; lets high-variance
metrics tolerate their own noise.
75% floor = empirical observation: even with 50 warmups, single-
host inter-run variance on bootstrap-cold queryd was
consistently 90-130% on this box. 75% catches >75%
regressions cleanly while ignoring known noise.
lib/metrics.sh: new proof_compute_mad helper computes MAD from a
file of one-number-per-line samples. Used for both regen (to write
the baseline.mad value) and diff (read from baseline).
Honest finding from this iteration's 3 back-to-back diff runs:
query_ms shows 90-130% delta from baseline consistently — not
random noise but a systematic 2x gap between regen-time and
steady-state. The regen captured a particularly fast moment;
steady-state is slower. Operator workflow: regenerate the
baseline at a known-representative state via
`bash tests/proof/run_proof.sh --mode performance --regenerate-baseline`
rather than expecting the harness to track a moving target.
The harness's value here is the EVIDENCE RECORD (every run captures
median+MAD+p95 plus all raw samples in raw/metrics/), not the gate.
Even false-positive REGRESSION skips give operators "this run was
20ms vs baseline 10ms" which is informative.
Sample counts also written into baseline.json under "samples" so a
future audit can verify the methodology that produced the values.
Verified across 3 back-to-back runs:
ingest_rows_per_sec PASS (delta within 75%, mostly < 10%)
vectors_per_sec_add PASS
search_ms PASS
rss_* PASS
query_ms REGRESSION flagged (130/100/90%) — known
systematic gap, not bug
Closes the "40% noise floor" follow-up from Phase E FINAL_REPORT.
Honest about limitations: hard regression gating on a busy single-
host setup needs either much bigger sample counts (n≥100), longer
warmup, or moving to a dedicated benchmark host. Documented inline.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
0d18ffa780 |
ADR-003: inter-service auth posture — Bearer + IP allowlist
Locks in the auth model that R-001 + R-007 will be retrofitted
against. Doc-only — wiring deferred to Sprint 1 when the first
non-loopback binding is needed.
Decision: Bearer token (from secrets-go.toml [auth] section) + IP
allowlist (CIDR list). Both layers required when auth is on; empty
token = G0 dev no-op. /health exempt.
Implementation shape (when it lands):
- internal/shared/auth.go middleware: one chi r.Use line per binary
- shared.Run gates: refuses non-loopback bind without configured token
- subtle.ConstantTimeCompare for token equality (timing-safe)
Alternatives considered + rejected:
mTLS — too heavy for single-machine inter-service traffic
JWT — buys nothing over Bearer without external IdP
IP-only — one stolen IP entry = full access; no defense depth
OAuth2 — no external IdP commitment in G0-G3 timeline
What this doesn't do:
- Doesn't implement (code lands Sprint 1)
- Doesn't break G0 dev (empty token = middleware no-op)
- Doesn't address gateway→end-user auth (different ADR shape)
Closes the design-decision blocker for R-001 and R-007. Wiring
ticket: Sprint 1 backlog story S1.2.
Also lifts ADR-002 (storaged per-prefix PUT cap) into the doc —
it was implemented in
|
||
|
rootandClaude Opus 4.7
|
423a3817c5 |
D: storaged per-prefix PUT cap — vectord _vectors/ → 4 GiB
Closes the documented 500K-test gap (memory project_golang_lakehouse:
"storaged 256 MiB PUT cap blocks single-file LHV1 persistence above
~150K vectors at d=768"). Vectord persistence under "_vectors/" now
gets a 4 GiB cap; everything else (parquets, manifests, ingest)
keeps the 256 MiB default.
Why per-prefix and not "raise globally":
- 256 MiB cap is a real DoS protection — runaway clients can't
drain the daemon. Raising it for ALL traffic would expand the
attack surface for routine paths that have no need.
- Per-prefix preserves existing protection while opening the one
documented production-scale path.
Why not split LHV1 across multiple keys (the alternative):
- G1P shipped a single-Put framed format SPECIFICALLY to eliminate
the torn-write class (memory: "Single Put eliminates the torn-
write class that the 3-way convergent scrum finding identified").
- Multi-key LHV1 would re-introduce the half-saved-state failure
mode we just paid to fix. Streaming via existing manager.Uploader
is the better architectural answer.
Why not bump the cap operationally via env/config:
- Future operator-driven cap can drop in cleanly via the
maxPutBytesFor function. Started with hardcoded 4 GiB to keep
this commit small; config knob is a follow-up if production
workloads diverge from the documented 500K-vector ceiling.
manager.Uploader is already streaming-multipart on the outbound
S3 side; the inbound MaxBytesReader cap is a safety gate, not a
memory bottleneck. So raising it for vectord just lets the
existing streaming path actually flow, without introducing new
memory pressure (4-slot semaphore × 4 GiB worst case = 16 GiB
only if all slots simultaneously max out — vanishingly unlikely).
Implementation:
cmd/storaged/main.go:
new constant maxPutBytesVectors = 4 GiB (covers >700K vectors @ d=768)
new constant vectorsPrefix = "_vectors/" (synced with vectord.VectorPrefix)
new function maxPutBytesFor(key) → cap-by-prefix
handlePut: ContentLength check + MaxBytesReader use the per-key cap
cmd/storaged/main_test.go (3 new test funcs):
TestMaxPutBytesFor: 7 cases incl. nested prefix, substring-but-not-
prefix, empty key, parquet/manifest paths.
TestVectorPrefixSyncWithVectord: regression test that asserts
vectorsPrefix == vectord.VectorPrefix. A future rename surfaces
here instead of silently bypassing the larger cap.
TestVectorCapAccommodates500KStaffingTest: bounds the cap above
the documented production workload (~700 MiB conservative).
Verified:
go test ./cmd/storaged/ — all green (was 1 func, now 4)
just verify — 9 smokes still pass · 32s wall
just proof contract — 53/0/1 unchanged
Out of scope for this commit (deserves its own):
- Heavy integration smoke: 200K dim=768 synthetic vectors → ~700
MiB LHV1 → kill+restart vectord → recall=1. ~5-10 min wall;
follow-up if you want production-scale persistence verified
end-to-end. Unit tests + existing g1p_smoke cover the wiring.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
6af0520ed2 |
A: fail-loud on non-loopback bind — closes worst case of R-001
shared.Run now refuses to bind a non-loopback address unless the
LH_<SERVICE>_ALLOW_NONLOOPBACK=1 env is set. Single change covers
all 7 binaries via the existing Run call site; no per-binary
wiring needed.
Closes the accidental-0.0.0.0 deploy attack surface for R-001:
queryd /sql is RCE-equivalent off loopback (DuckDB has filesystem
read + COPY TO + read_text), but the gate applies to every binary
uniformly so the same posture covers vectord (mutation routes),
catalogd (manifest writes), and the others.
What passes the gate:
127.0.0.1:port, 127.x.y.z:port (full /8), [::1]:port,
localhost:port, OR explicit env LH_<SVC>_ALLOW_NONLOOPBACK=1
What fail-louds:
0.0.0.0:port, [::]:port, :port (all interfaces),
any non-loopback IP, any non-localhost hostname,
unparseable shapes ("", "no port", garbage)
Override env is strict equality "1" — typos like "true"/"yes" do NOT
trigger it, so a future operator can't accidentally expose by typing
the wrong value. Override fires log a structured warn so the choice
is auditable in production.
Error message cites the env name AND R-001 by name so operators see
the fix path without grepping:
"refusing non-loopback bind \"0.0.0.0:3214\" for \"queryd\"
(set LH_QUERYD_ALLOW_NONLOOPBACK=1 to override; see audit R-001)"
internal/shared/bind.go — requireLoopbackOrOverride + isLoopbackAddr
internal/shared/bind_test.go — 7 test funcs incl. table-driven
IPv4/IPv6/hostname coverage and
per-service env isolation
internal/shared/server.go — 1-line gate in Run before listen
Verified:
go test -short ./internal/shared/ — all green (was 14 funcs, now 21)
just verify — vet + test + 9 smokes still 33s
Doesn't address R-001's full attack surface (any reachable port can
issue arbitrary SQL); ADR-003 + Bearer-token middleware is the
follow-up. This commit makes the implicit "localhost-only is the auth
layer" guarantee explicit and un-bypassable without explicit env.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
125e1c80b9 |
tests: close R-002 / R-003 / R-008 — internal/shared, storeclient, queryd/db.go
Audit-driven follow-up to the Rust scrum review on the 3 untested
HIGH-risk packages. Both the audit (reports/scrum/risk-register.md)
and the scrum (tests/real-world/runs/scrum_mojxb5bw/) independently
flagged these files as the highest-leverage missing test coverage.
internal/shared/server_test.go — 8 test funcs
newListener: valid addr, invalid addr (non-numeric port, port
out of range, port-already-in-use surfacing as net.OpError).
Empty-addr-is-valid: documents the net.Listen quirk that "" binds
an OS-picked port — future readers don't need to relitigate.
HealthResponse marshal: JSON shape stable, round-trip clean.
/health handler reconstructed via httptest.Server: status 200,
Content-Type application/json, body fields stable.
RegisterRoutes callback: contract verified (callback is invoked
with a real chi.Router, mounted route reachable end-to-end).
Run bind-failure surface: synchronous error, not a goroutine swallow
— the contract Run depends on per the race-safe-startup comment.
internal/shared/config_test.go — 6 test funcs
DefaultConfig G0 port pinning: every binary's default bind locked
in (3110/3211-3216) so a refactor can't silently flip a port.
LoadConfig empty path: returns DefaultConfig, no error.
LoadConfig missing file: returns DefaultConfig, logs warn (the warn
line shows up in test output, captured-but-not-asserted).
LoadConfig valid TOML: partial overrides land, unspecified sections
keep defaults (TOML decoder leave-alone behavior).
LoadConfig invalid TOML: returns wrapped 'parse config' error.
LoadConfig unreadable file: skipped under root (root reads 0000);
captures the read-error wrap path for non-root contexts.
internal/storeclient/client_test.go — 14 test funcs
safeKey table-driven: plain segments, single slash, empty, trailing
slash, space (→ %20), apostrophe (→ %27), unicode (→ %C3%A9),
deep nesting. Locks URL-escape contract per scrum suggestion.
recordingServer helper backs Put/Get/Delete/List against
httptest.Server: verifies method, path, body bytes round-trip.
ErrKeyNotFound on 404 (errors.Is round-trip).
Non-OK status wraps body preview into the error chain.
Delete accepts both 200 and 204 (S3 vs compatible-store quirk).
List parses JSON shape and surfaces query-string prefix.
Context cancellation propagates through Put as context.Canceled.
internal/queryd/db_test.go — 5 test funcs (with subtests)
sqlEscape table-driven: 8 cases including empty, all-quotes,
nested apostrophes (the case from the scrum suggestion).
redactCreds table-driven: 6 cases — both keys, single keys,
empty, multi-occurrence, placeholder-collision (lossy but safe).
buildBootstrap statement order: INSTALL → LOAD → CREATE SECRET.
buildBootstrap endpoint schemes: http strips + USE_SSL false,
https keeps SSL true, no-scheme defaults SSL true (prod ambient).
buildBootstrap URL_STYLE: 'path' vs 'vhost' branch.
buildBootstrap escapes credential quotes: future SSO-token-with-
apostrophe doesn't break out of the SQL string literal — the
belt holds when the suspenders snap.
Real finding caught by my own test:
net.Listen("tcp", "") succeeds (OS-picked port) — captured as
TestNewListener_EmptyAddrIsValid so the quirk is documented.
Verified:
go test -short ./... — every internal/ package now has tests
(no more 'no test files' lines for shared/storeclient).
just verify — vet + test + 9 smokes green in 33s.
just proof contract — 53/0/1 green (no harness regression).
Closes:
R-002 internal/shared zero tests HIGH
R-003 internal/storeclient zero tests HIGH
R-008 queryd/db.go untested MED (sqlEscape, redactCreds,
CREATE SECRET formation)
Composite scrum score should move from 43 → ~46 / 60 — the three
HIGH/MED risks closed, internal/shared and internal/storeclient
become "tested + load-bearing" instead of "untested + load-bearing."
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
ff9823b871 |
scrum audit re-run: 35 → 43 / 60 after Phase A-E + S0.3
Re-runs the SCRUM.md framework against HEAD ( |
||
|
rootandClaude Opus 4.7
|
4840c10311 |
proof harness: fix queryd refresh-tick race in 04_query_correctness
Caught by the audit rerun: with cache-warm binaries, 04 fires its
first SELECT faster than queryd's 500ms refresh tick — Q1 returned
400 ("table not found") even though 03_ingest had registered the
manifest. Subsequent queries (after the next tick) succeeded.
This is an eventual-consistency wait, not a retry — queryd's
contract is that views appear within one tick of catalogd having the
manifest. Production code does not need changing.
Added to lib/http.sh:
proof_wait_for_sql <budget_sec> <sql>
polls a SQL probe until it returns 200 or budget elapses; emits
no evidence (test setup, not a claim).
Used in 04_query_correctness:
Wait up to 5s for queryd to have the view before running the 5
SQL assertions. Skip-with-loud-reason if the view never appears.
Verified: integration mode back to 104 pass / 0 fail / 1 skip after
fix. The skip is the unchanged GOLAKE-085 informational record.
This is exactly the kind of finding the harness was designed to
surface — the regression existed in the codebase the moment Phase D
shipped, but only fired when the next compare run hit cache-warm
timing. Without the harness, it would have surfaced on a CI run
weeks from now and been hard to bisect.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
4bb6548cbc |
proof harness Phase E: FINAL_REPORT.md answers the 9 mandated questions
Per docs/TEST_PROOF_SCOPE.md, this is the closing deliverable for the
proof harness: a single document that names what's proven, what's
partially proven, what failed, what was skipped and why, what evidence
exists for each, what bottlenecks were measured, what contract drift
was found, what refactor risks remain, and what to fix first.
Per-run report dirs (tests/proof/reports/proof-<ts>/) keep their
existing summary.md + summary.json + raw/ structure — they are the
replayable evidence chain. FINAL_REPORT.md is the stable, repo-tracked
synthesis pointing at them.
Headline findings (no surprises — harness behaves as designed):
- 24 claims encoded; 22 fully proven, 1 informational (GOLAKE-085
duplicate vector ID, contract not yet specified), 0 failed.
- 4 contract-drift findings recorded as canonical: vectord add
body field is `items` not `vectors`, search response is `results`
not `hits`, index info is `length` not `count`, status codes
201/204 not 200. All caught during Phase B; all now pinned by the
harness.
- Performance baseline shows queryd as the largest RSS (69 MiB,
DuckDB process); single-sample noise floor is ~40% — tightening
to multi-sample medians is a documented Sprint follow-up.
- HIGH-risk audit findings (R-001 queryd /sql, R-002/R-003 untested
shared+storeclient) are NOT closed by the harness — it's a
multiplier, not a replacement for unit tests + auth posture.
The proof harness is complete. 11 cases · 3 modes · 168 assertions
peak across all tiers · ~22s total wall (contract+integration+perf).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
175ad59cb3 |
proof harness Phase D: performance baseline · 1000-row ingest, p50/p95
GOLAKE-100. First run writes tests/proof/baseline.json; subsequent
runs diff against it. >10% regression emits a SKIP with REGRESSION
detail (not a fail — perf claim is required:false in claims.yaml so
the gate stays green; the human summary tells the regression story
honestly). Skip-with-loud-reason if any earlier case in the run
failed, per spec "performance only after contract+integration pass."
Workload (deterministic, repeatable):
ingest 1000-row CSV (5 roles × 5 cities × seeded scores) → /v1/ingest
query SELECT count(*) ×20 against the just-ingested dataset
vector add 200 dim=4 vectors with formulaic content (no Ollama)
search ×20 against the perf index with a fixed query vector
RSS per-service post-workload sample via /proc/<pid>/status
Recorded metrics:
ingest_rows_per_sec, query_p50_ms, query_p95_ms,
vectors_per_sec_add, search_p50_ms, search_p95_ms,
rss_{storaged,catalogd,ingestd,queryd,vectord,embedd,gateway}_mb
baseline.json on this box (committed):
25000 rows/sec ingest · 17ms p50 / 24ms p95 query
6250 vectors/sec add · 8ms p50 / 20ms p95 search
queryd 69 MiB · vectord 14 MiB · others 11-29 MiB
Honest measurement-design finding from the very first compare run:
back-to-back runs surfaced -41% ingest and +29% query p50 — pure
disk-cache + queryd-cold-start noise. Single-sample baselines have
real noise floor ≈40%. Recorded as REGRESSION skips so the human
summary surfaces it, not a code regression. Tightening the threshold
or moving to multi-sample medians is a Phase E recommendation.
Verified end-to-end:
just proof contract — 53 pass · 1 skip · ~4s
just proof integration — 104 pass · 1 skip · ~8s
just proof performance — 110 pass · 3 skip · ~10s
just verify — 9 smokes still green · 29s
All 11 cases (4 contract + 6 integration + 1 performance) deterministic
end-to-end. Phase E (final report against the 9 mandated questions)
is the last piece.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
1313eb2173 |
proof harness Phase C: 6 integration cases · 104/0/1 green
Adds the integration tier — full chain CSV→Parquet→SQL and full
text→embed→vector→search. All 10 cases (4 contract + 6 integration)
end-to-end deterministic; 8s wall total.
Cases added:
01_storage_roundtrip.sh
GOLAKE-010-012. PUT 1KiB → GET sha256-equal → LIST contains key
→ DELETE 200/204 → GET 404. Deterministic key under
proof/<case_id>/ so concurrent runs don't collide.
02_catalog_manifest.sh
GOLAKE-020-022. Fresh register existing=false → manifest read
matches → list contains dataset_id → idempotent re-register
existing=true with stable dataset_id → schema-drift register
409 (the ADR-020 contract). Per-run unique name via
PROOF_RUN_ID so existing=false is meaningful.
03_ingest_csv_to_parquet.sh
GOLAKE-030. workers.csv (5 rows) via /v1/ingest multipart →
parquet object on storaged → catalog manifest with row_count=5.
Verifies content-addressed key shape (datasets/<n>/<fp>.parquet).
04_query_correctness.sh
GOLAKE-040. The 5 SQL assertions from fixtures/expected/queries.json
against the workers fixture: count=5, Chicago=2, max=95,
safety→Barbara, Houston avg=89.5. Iterates the YAML claims, runs
each query, compares response columns to expected values.
06_vector_add_search.sh integration extension
GOLAKE-051. text → /v1/embed (4 docs from fixtures/text/docs.txt)
→ vectord add → search by query embedding. Top-1 ID per query
asserted against fixtures/expected/rankings.json. First run (or
--regenerate-rankings) writes the fixture and emits a skip with
explicit reason; subsequent runs assert against it.
07_vector_persistence_restart.sh
GOLAKE-070. add 4 unit-basis vectors → search → record top-1
distance → SIGTERM vectord → restart with the same --config →
poll /health for 8s → search again → top-1 ID and distance match
bit-identically. Skips with reason if vectord PID can't be found
or post-restart bind times out.
Two harness improvements landed alongside:
run_proof.sh writes a temp lakehouse_proof.toml with
refresh_every="500ms" override and passes --config to all booted
binaries. Production default is 30s; 04_query_correctness needs
queryd to pick up the new view within a tick. Production config
unchanged.
cleanup() now pgreps for any orphan bin/<svc> processes (anchored
to start-of-argv per memory feedback_pkill_scope.md) so a case
that restarts a service mid-run still gets cleaned up.
lib/http.sh adds proof_call(case_id, probe, method, url, args...)
— escape hatch for cases that need raw curl args (multipart -F,
custom headers). Used by 03_ingest for the multipart upload that
conflicts with proof_post's --data + Content-Type defaults.
lib/env.sh exports PROOF_RUN_ID — short unique id derived from the
report directory timestamp. Used by 02 and 07 for fresh-each-run
state isolation.
Two real findings recorded as evidence (no code changes):
- rankings.json fixture pinned: 4 queries → 4 distinct top-1 docs
via nomic-embed-text. A model swap that changes ranking now
fails the harness loudly; --regenerate-rankings is the override.
- vectord persistence kill+restart preserves top-1 distance
bit-identically — the LHV1 single-Put framed format from
G1P round-trips exactly through Save/Load.
Verified end-to-end:
just proof contract — 53 pass (4 cases)
just proof integration — 104 pass (10 cases) · 8s wall
just verify — 9 smokes still green · 33s wall
Phase D (performance baseline) lands next: 10_perf_baseline measures
rows/sec ingest, vectors/sec add, p50/p95 query+search latency, RSS,
CPU. First run writes tests/proof/baseline.json; later runs diff
against it.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
6d18394416 |
proof harness Phase B: 4 contract cases · 53/0/1 green
Added the contract tier above 00_health canary. All 5 contract cases
now cover GOLAKE-001-003, 050, 060-061, 080-085 — 53 assertions pass,
1 informational skip, 0 fail. Wall: 4s end-to-end (cached binaries).
Cases:
05_embedding_contract.sh
GOLAKE-050. POST /v1/embed with one short text → asserts dim=768,
one vector returned, vector length matches dimension, sum of
squared elements > 0 (proxy for non-zero), response.model echoed.
Skips with explicit reason if Ollama is unreachable (502 from
embedd) — per spec hard rule "skipped tests do not appear as
passed."
06_vector_add_search.sh
GOLAKE-060 + GOLAKE-061. Synthetic dim=4 unit basis vectors.
Create index → add 3 vectors → get-index returns length=3 →
search([1,0,0,0],k=3) returns v1 at rank 1 with distance < 0.001.
Cleanup with DELETE. No embedd dependency — pure contract layer.
08_gateway_contracts.sh
GOLAKE-003. For each /v1/* route, asserts gateway and direct
upstream return identical status AND identical response body
(sha256 match). Confirms gateway is a proxy not a transformer.
Status passthrough verified on both 200 path (storage/list,
catalog/list) and 4xx path (sql empty body → 400 from queryd).
09_failure_modes.sh
GOLAKE-080..085. Six failure-mode contracts:
080 malformed JSON → 4xx on catalog/ingest/sql/embed
081 missing required field → 4xx on catalog/vectors/embed
082 bad SQL → 4xx with non-empty error body
083 vector dim mismatch → 4xx
084 missing storage object → 404
085 duplicate vector ID → INFORMATIONAL (spec says required:false)
first/second statuses recorded as evidence; contract decided
later from the recorded record.
Two new lib helpers in lib/assert.sh:
proof_assert_status_in <id> <claim> "200 201 204" <probe>
pass if status is in the space-separated list. Used for
delete-returns-200-or-204 case where vectord returns 204.
proof_assert_status_4xx <id> <claim> <probe>
pass if status in [400, 500). Used for failure modes where the
specific 4xx code may vary (400 vs 422 vs 409). Records actual
code as evidence.
Two real contract findings recorded by the harness during build:
- vectord add expects {"items": [...]}, not {"vectors": [...]}.
My initial test sent the wrong field; would have masked the bug
forever in CI. The harness caught it via the assertion failure.
- vectord create returns 201 Created, delete returns 204 No Content.
Documented in the test fixtures as canonical.
Regression: just verify wall 33s, vet + test + 9 smokes still green.
Phase C (integration) lands next: 01_storage_roundtrip, 02_catalog_manifest,
03_ingest_csv_to_parquet, 04_query_correctness, 05/06 integration extends,
07_vector_persistence_restart.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
a81291e38c |
proof harness Phase A: scaffolding + canary case green
Per docs/TEST_PROOF_SCOPE.md, building the claims-verification tier
above the smoke chain. This commit lays the scaffolding and proves
the orchestrator end-to-end with one canary case (00_health).
What landed:
tests/proof/
README.md how to read a report, layout, modes
claims.yaml 24 claims enumerated (GOLAKE-001..100)
run_proof.sh orchestrator with --mode {contract|integration|performance}
and --no-bootstrap / --regenerate-{rankings,baseline}
lib/
env.sh service URLs, report dir, mode, git context
http.sh curl wrappers writing per-probe JSON + body + headers
assert.sh proof_assert_{eq,ne,contains,lt,gt,status,json_eq} +
proof_skip — each emits one JSONL record per call
metrics.sh start/stop timers, value capture, RSS sampling,
percentile compute (for Phase D)
cases/
00_health.sh canary — gateway + 6 services /health → 200,
body identifies service, latency < 500ms (21 assertions)
fixtures/
csv/workers.csv spec's 5-row deterministic CSV
text/docs.txt 4 deterministic vector docs
expected/queries.json expected results for the 5 SQL assertions
Wired into the task runner:
just proof contract # canary only this commit
just proof integration # Phase C
just proof performance # Phase D
.gitignore: /tests/proof/reports/* with !.gitkeep — same pattern as
reports/scrum/_evidence/. Per-run output is a runtime artifact.
Specs landed alongside (J's drops):
docs/TEST_PROOF_SCOPE.md the harness contract this implements
docs/CLAUDE_REFACTOR_GUARDRAILS.md process discipline this harness obeys
Verified end-to-end (cached binaries):
just proof contract wall < 2s, 21 pass / 0 fail / 0 skip
just verify wall 31s, vet + test + 9 smokes still green
Two bugs fixed during canary run, both in run_proof.sh aggregation:
- grep -c exits 1 on zero matches; the `|| echo 0` form concatenated
"0\n0" and broke jq --argjson + integer comparison. Fixed via a
_count helper that captures count-or-zero cleanly.
- per-case table iterated case scripts (filename-based) but cases
write evidence under CASE_ID. Switched to JSONL-file iteration so
multi-case scripts work and the mapping is faithful.
Phase B (contract cases) lands next: 05_embedding, 06_vector_add,
08_gateway_contracts, 09_failure_modes. Each sourcing the same lib
helpers and writing to the same report shape.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
e31638204d |
S0.3: just verify + pre-push hook gates the smoke chain
Sprint 0 / R-004 / GATE-0.4 — the 9-smoke chain is no longer
documentation only. One command (`just verify`) runs vet + tests +
all 9 smokes; pre-push hook calls it; a regression cannot leave
this machine without explicit --no-verify override.
Recipes:
just verify full gate (33s wall on this box)
just smoke <day> single smoke (d1..d6, g1, g1p, g2)
just smoke-all all 9 smokes only
just doctor dep probe with structured output
(--json for CI / pre-push)
just install-hooks install .git/hooks/pre-push
just fmt|vet|test|build|clean
scripts/doctor.sh probes Go ≥1.25, gcc, MinIO at :9000 with bucket
lakehouse-go-primary, Ollama at :11434 with nomic-embed-text loaded,
/etc/lakehouse/secrets-go.toml with [s3.primary]. Each missing dep
prints its install fix command. JSON mode emits the same shape for
CI / pre-push consumers.
README updated with the task-runner section + just install-hooks
on cold-start. Hooks live in .git/hooks/ (untracked); install
recipe recreates them on a fresh clone.
PATH note: justfile prepends /usr/local/go/bin so recipes find Go
without depending on the parent shell's PATH (ADR-001 §1.x lives
go there).
Verified: just verify exits 0 in 33s wall (vet ~0.1s + test ~0.1s +
9 smokes deterministic per audit baseline). Pre-push hook installed
and bash -n clean.
Closes audit risk R-004 (smokes not gated).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
91edd43164 |
scrum audit: 5 reports under reports/scrum/ · score 35/60
Adapts docs/SCRUM.md framework (originally written for the matrix-agent-validated repo) to the Go rewrite. Five deliverables: golang-lakehouse-scrum-test.md top-line + scoring + verdict risk-register.md 12 findings, R-001..R-012 claim-coverage-table.md claim/test/risk for Sprint 2 sprint-backlog.md 5 sprints, ~2 weeks of work acceptance-gates.md DoD as runnable commands Every claim cites file:line, command output, or "missing evidence." Smoke chain ran clean (33s wall, all 9 PASS) and is captured in reports/scrum/_evidence/smoke_chain.log (gitignored — runtime artifact). Scoring: Reproducibility 7/10 9 smokes deterministic, no just/CI gate Test Coverage 6/10 internal/ packages tested, 6/7 cmd/ aren't Trust Boundary 7/10 escapes ok, zero auth, /sql is RCE-eq off-loopback Memory Correctness 3/10 pathway/playbook/observer not yet ported Deployment Readiness 4/10 no REPLICATION, no env template, no systemd Maintainability 8/10 no god-files, 7 lean binaries, ADRs current Top three risks: R-001 HIGH queryd /sql + DuckDB + non-loopback bind = RCE-equivalent R-002 HIGH internal/shared (server.go + config.go) zero tests R-003 HIGH internal/storeclient zero tests, used by 2 services R-004 MED 9-smoke chain green but not gated (no justfile/hook) The audit is the work; refactors come after. Sprint 0 owns coverage + CI gating; Sprint 1 owns trust-boundary decisions; Sprints 2-3 are mostly design-bar work for unbuilt agent components. .gitignore exception: /reports/* + !/reports/scrum/ keeps reports/ a runtime-artifact directory while exposing reports/scrum/ as tracked documentation. Mirrors the pattern future audit passes will land in. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
1f700e731d |
Staffing scale test: full 500K through gateway → embedd → vectord pipeline
scripts/staffing_500k/main.go: driver that reads workers_500k.csv,
embeds combined-text per worker via /v1/embed, adds to vectord index
"workers_500k", runs canonical staffing queries against the populated
index. Reproducible end-to-end test of the staffing co-pilot pipeline
at production scale.
Run results (2026-04-29 ~02:30):
500,000 vectors ingested in 35m 36s (~234/sec avg)
vectord peak RSS 4.5 GB (~9 KB/vector incl. HNSW graph)
Query latency: embed 40-59ms + search 1-3ms = ~50ms end-to-end
GPU avg ~65% (Ollama not the bottleneck — vectord Add is)
Semantic recall on canonical queries:
"electrician with industrial wiring": top 2 are literal Electricians (d=0.30)
"CNC operator with first article": Assembler / Quality Techs (adjacent, d=0.24)
"forklift driver OSHA-30": warehouse roles (d=0.33)
"warehouse picker night shift bilingual": Material Handlers (d=0.31)
"dental hygienist": Production Workers at d=0.49+ — correctly
LOW-similarity, signals "no dental hygienists in this manufacturing
dataset" rather than hallucinating a fake match.
Documented gaps:
- storaged's 256 MiB PUT cap blocks single-file LHV1 persistence
above ~150K vectors at d=768. Test ran with persistence disabled.
- vectord Add is RWMutex-serialized — with GPU at 65% util this is
the throughput cap. Concurrent Adds would be 2-3x faster but
require careful audit of coder/hnsw thread-safety (G1 scrum
documented two known quirks).
PHASE_G0_KICKOFF.md gains a "Staffing scale test" section with full
metrics + the gaps-surfaced list. The architectural payoff is real:
six binaries, one HTTP route, ~50ms from text query to top-K
semantically-relevant workers across 500K records.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
0cb29cda15 |
docs: README + PHASE_G0_KICKOFF reflect post-G0 state (G1, G1P, G2)
README was stuck on "Pre-Phase G0, implementation has not started" while we shipped through G2. Updated to reflect the current 7-binary service inventory, the 9 acceptance smokes, the cold-start deps (MinIO bucket, Ollama with nomic-embed-text, secrets-go.toml). PHASE_G0_KICKOFF gains a "Post-G0 work" pointer at the end — brief table mapping each G1+/G2 commit to its smoke + scrum-fix count. Full per-day detail stays in commit messages and the project memory file. No code changes. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
9ee7fc5550 |
G2: embedd — text → vector via Ollama · 2 scrum fixes
Bridges the missing piece for the staffing co-pilot: text inputs to
vectord-shaped vectors. Standalone cmd/embedd on :3216 fronted by
gateway at /v1/embed. Pluggable embed.Provider interface (G2 ships
Ollama; OpenAI/Voyage swap in via the same interface in G3+).
Wire format:
POST /v1/embed {"texts":[...], "model":"..."} // model optional
→ 200 {"model","dimension","vectors":[[...]]}
Default model: nomic-embed-text (768-d). Ollama returns float64;
provider converts to float32 at the boundary so vectors flow through
vectord/HNSW without re-conversion.
Acceptance smoke 5/5 PASS — including the architectural payoff:
end-to-end embed → vectord add → search by re-embedded text returns
recall=1 at distance 5.96e-8 (float32 precision noise on identical
unit vectors). The staffing co-pilot pipeline (text → vector →
similarity search) is now functional end-to-end.
All 9 smokes (D1-D6 + G1 + G1P + G2) PASS deterministically.
Cross-lineage scrum on shipped code:
- Opus 4.7 (opencode): 0 BLOCK + 4 WARN + 3 INFO
- Kimi K2-0905 (openrouter): 0 BLOCK + 2 WARN + 1 INFO
- Qwen3-coder (openrouter): "No BLOCKs" (3 tokens)
Fixed (2 — 1 convergent + 1 single-reviewer):
C1 (Opus + Kimi convergent WARN): per-text 60s timeout × N-text
batch was up to N×60s with no batch-level cap. One stuck Ollama
call would stall the whole handler indefinitely. Fix:
context.WithTimeout(r.Context(), 60s) wraps the entire batch.
O-W3 (Opus WARN): empty strings in texts went to Ollama unchecked,
producing version-dependent garbage. Fix: reject "" with 400 at
the handler boundary so callers get a deterministic answer
instead of an upstream-conditional 502.
Deferred (4): drainAndClose 64KiB cap (matches G0 pattern), no
concurrency limit on /embed (single-tenant G2), missing Accept
header (exotic-proxy concern), MaxBytesError string-match
redundancy (paranoia layer kept consistent across codebase).
Zero false positives this round — Qwen returned 3 tokens "No BLOCKs"
and the other two reviewers' findings were all real.
Setup confirmed: Ollama 0.21.0 on :11434 with nomic-embed-text loaded.
Per-text /api/embeddings used (forward-compat with 0.21+); newer
0.4+ /api/embed batch endpoint can swap in via the Provider interface.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
8b92518d21 |
G1P: vectord persistence to storaged + scrum (3 fixes incl. 3-way convergent)
Adds optional persistence to vectord (G1's HNSW vector search). Single-
file framed format per index — eliminates the torn-write class that
the 3-way convergent scrum finding identified:
_vectors/<name>.lhv1 — single binary blob:
[4 bytes magic "LHV1"]
[4 bytes envelope_len uint32 BE]
[envelope bytes — JSON params + metadata + version]
[graph bytes — raw hnsw.Graph.Export]
Pre-extraction: internal/catalogd/store_client.go → internal/storeclient/
shared package, since both catalogd and vectord need it. Same pattern as
the pre-D5 catalogclient extraction.
Optional via [vectord].storaged_url config (empty = ephemeral mode).
On startup: List + Load each persisted index. After Create / batch Add /
DELETE: Save (or Delete from storaged). Save failures are logged-not-
fatal — in-memory state is the source of truth in flight.
Acceptance smoke G1P 8/8 PASS — kill+restart preserves state, post-
restart search returns dist=0 (graph round-trips exactly), DELETE
removes the file, post-delete restart shows count=0.
All 8 smokes (D1-D6 + G1 + G1P) PASS deterministically. The g1_smoke
gained scripts/g1_smoke.toml that disables persistence so the
in-memory API test stays decoupled from any rehydrate-from-storaged
state contamination.
Cross-lineage scrum on shipped code:
- Opus 4.7 (opencode): 1 BLOCK + 5 WARN + 3 INFO
- Kimi K2-0905 (openrouter): 1 BLOCK + 2 WARN
- Qwen3-coder (openrouter): 2 BLOCK + 2 WARN + 1 INFO
Fixed (3 — 1 convergent + 2 single-reviewer):
C1 (Opus + Kimi + Qwen 3-WAY CONVERGENT WARN): Save was non-atomic
across two PUTs — envelope-succeeds + graph-fails left a half-
saved index that passed the "both present" List filter and
silently mismatched metadata against vectors on Load.
Fix: collapse to single framed file (no torn-write window
possible).
O-B1 (Opus BLOCK): isNotFound substring-matched "key not found"
against the wrapped error message — brittle, any 5xx body
containing that text would silently misclassify as missing.
Fix: errors.Is(err, storeclient.ErrKeyNotFound).
O-I3 (Opus INFO): handleAdd pre-validation only covered id+dim;
NaN/Inf/zero-norm could still fail mid-batch leaving partial
commits. Fix: extend pre-validation to call ValidateVector
(newly exported) per item before any commit.
Dismissed (3 false positives):
K-B1 + Q-B1 ("safeKey double-escapes %2F segments") — false
convergent. Wire-protocol escape is decoded by storaged's chi
router on the way in; on-disk key is the original literal.
%2F round-trips correctly through PathEscape → URL → chi decode
→ S3 key.
Q-B2 ("List vulnerable to race conditions") — vectord is single-
process; no concurrent Save against List in the same vectord.
Deferred (3): rehydrate per-index timeout (G2+ multi-index scale),
saveAfter request ctx (matches G0 timeout deferral), Encode RLock
during slow writer (documented as buffer-only API).
The C1 finding is the strongest signal of the cross-lineage filter:
three independent reviewers all flagged the same torn-write hazard.
Single-file framing eliminates the class — there's now no Persistor
state where envelope and graph can disagree.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
b8c072cf0b |
G1: vectord — HNSW vector search via coder/hnsw · 6 scrum fixes applied
First G1+ piece. Standalone vectord service with in-memory HNSW
indexes keyed by string IDs and optional opaque JSON metadata.
Wraps github.com/coder/hnsw v0.6.1 (pure Go, no cgo). New port
:3215 with /v1/vectors/* routed through gateway.
API:
POST /v1/vectors/index create
GET /v1/vectors/index list
GET /v1/vectors/index/{name} get info
DELETE /v1/vectors/index/{name}
POST /v1/vectors/index/{name}/add (batch)
POST /v1/vectors/index/{name}/search
Acceptance smoke 7/7 PASS — including recall=1 on inserted vector
w-042 (cosine distance 5.96e-8, float32 precision noise), 200-
vector batch round-trip, dim mismatch → 400, missing index → 404,
duplicate create → 409.
Two upstream library quirks worked around in the wrapper:
1. coder/hnsw.Add panics with "node not added" on re-adding an
existing key (length-invariant fires because internal
delete+re-add doesn't change Len). Pre-Delete fixes for n>1.
2. Delete of the LAST node leaves layers[0] non-empty but
entryless; next Add SIGSEGVs in Dims(). Workaround: when
re-adding to a 1-node graph, recreate the underlying graph
fresh via resetGraphLocked().
Cross-lineage scrum on shipped code:
- Opus 4.7 (opencode): 0 BLOCK + 4 WARN + 3 INFO
- Kimi K2-0905 (openrouter): 2 BLOCK + 2 WARN + 1 INFO
- Qwen3-coder (openrouter): "No BLOCKs" (4 tokens)
Fixed (4 real + 2 cleanup):
O-W1: Lookup returned the raw []float32 from coder/hnsw — caller
mutation would corrupt index. Now copies before return.
O-W3: NaN/Inf vectors poison HNSW (distance comparisons return
false for both < and >, breaking heap invariants). Zero-norm
under cosine produces NaN. Now validated at Add time.
K-B1: Re-adding with nil metadata silently cleared the existing
entry — JSON-omitted "metadata" field deserializes as nil,
making upsert non-idempotent. Now nil = "leave alone"; explicit
{} or Delete to clear.
O-W4: Batch Add with mid-batch failure left items 0..N-1
committed and item N rejected. Now pre-validates all IDs+dims
before any Add.
O-I1: jsonItoa hand-roll replaced with strconv.Itoa — no
measured allocation win.
O-I2: distanceFn re-resolved per Search → use stored i.g.Distance.
Dismissed (2 false positives):
K-B2 "MaxBytesReader applied after full read" — false, applied
BEFORE Decode in decodeJSON
K-W1 "Search distances under read lock might see invalidated
slices from concurrent Add" — false, RWMutex serializes
write-lock during Add against read-lock during Search
Deferred (3): HTTP server timeouts (consistent G0 punt),
Content-Type validation (internal service behind gateway), Lookup
dim assertion (in-memory state can't drift).
The K-B1 finding is worth pausing on: nil metadata on re-add is
the kind of API ergonomics bug only a code-reading reviewer
catches — smoke would never detect it because the smoke always
sends explicit metadata. Three lines changed in Add; the resulting
API matches what callers actually expect.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
d023b07b30 |
Real-scale validation post-G0: configurable ingest cap + workers_500k metrics
Validated G0 substrate against the production workers_500k.parquet dataset (18 cols × 500,000 rows). Findings + one applied fix: Finding #1 (FIXED): ingestd's hardcoded 256 MiB cap rejected the 500K CSV (344 MiB) with 413. Cap fired correctly, no OOM. Extracted to [ingestd].max_ingest_bytes config field; default 256 MiB, override per deployment for known-large workloads. With cap bumped to 512 MiB, 500K ingest succeeds in 3.12s with ingestd peak RSS 209 MiB. Finding #2 (deferred): ingestd doesn't release memory between ingests. Go runtime conservative; long-running daemon, fine. Finding #3: DuckDB-via-httpfs is healthy at 500K. GROUP BY 45ms, count(*) 24ms, AVG 47ms, schema introspection 25ms. Sub-linear scaling vs 100K — the s3:// read path is not a bottleneck. Finding #4: ADR-010 type inference correctly handled real staffing data. worker_id → BIGINT, numeric scores → DOUBLE, multi-line resume_text → VARCHAR. 1000-row sample sufficient. Finding #5: Go's encoding/csv handles RFC 4180 quoted-comma fields and multi-line quoted text without LazyQuotes — confirming the D4 scrum's dismissal of Qwen's BLOCK on this point. Net: substrate handles production-scale data with one config knob. No correctness issues, no OOMs, no silent type errors. All 6 G0 smokes still PASS after the cap-config change. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
rootandClaude Opus 4.7
|
b1d52306ad |
G0 D6: gateway reverse proxy fronting all 4 backing services · 2 scrum fixes · G0 COMPLETE
Last day of Phase G0. Gateway promotes the D1 stub endpoints into
real reverse-proxies on :3110 fronting storaged + catalogd + ingestd
+ queryd. /v1 prefix lives at the edge — internal services route on
/storage, /catalog, /ingest, /sql, with the prefix stripped by a
custom Director per Kimi K2's D1-plan finding.
Routes:
/v1/storage/* → storaged
/v1/catalog/* → catalogd
/v1/ingest → ingestd
/v1/sql → queryd
Acceptance smoke 6/6 PASS — every assertion goes through :3110, none
direct to backing services. Full ingest → storage → catalog → query
round-trip verified end-to-end. The smoke's "rows[0].name=Alice"
assertion is the architectural payoff: five binaries, six HTTP
routes, one round-trip through one edge.
Cross-lineage scrum on shipped code:
- Opus 4.7 (opencode): 1 BLOCK + 2 WARN + 2 INFO
- Kimi K2-0905 (openrouter): 1 BLOCK + 3 WARN + 1 INFO (3 false positives, all from one wrong TrimPrefix theory)
- Qwen3-coder (openrouter): 5 completion tokens — "No BLOCKs."
Fixed (2, both Opus single-reviewer):
O-BLOCK: Director path stripping fails if upstream URL has a
non-empty path. The default Director's singleJoiningSlash runs
BEFORE the custom code, so an upstream like http://host/api
produces /api/v1/storage/... after the join — then TrimPrefix("/v1")
is a no-op because the string starts with /api. Fix: strip /v1
BEFORE calling origDirector. New TestProxy_SubPathUpstream regression
locks this in. Today: bare-host URLs only, dormant — but moving
gateway behind a sub-path in prod would have silently 404'd.
O-WARN2: url.Parse is permissive — typo "127.0.0.1:3211" (no scheme)
parses fine, produces empty Host, every request 502s. mustParseUpstream
fail-fast at startup with a clear message naming the offending
config field.
Dismissed (3, all Kimi, same false TrimPrefix theory):
K-BLOCK "TrimPrefix loops forever on //v1storage" — false, single
check-and-trim, no loop
K-WARN "no upper bound on repeated // removal" — same false theory
K-WARN "goroutines leak if upstream parse fails while binaries
running" — confused scope; binaries are separate OS processes
launched by the smoke script
D1 smoke updated (post-D6): the 501 stub probes are gone (gateway no
longer stubs /v1/ingest and /v1/sql). Replaced with proxy probes that
verify gateway forwards malformed requests to ingestd and queryd. Launch
order changed from parallel to dep-ordered (storaged → catalogd →
ingestd → queryd → gateway) since catalogd's rehydrate now needs
storaged, queryd's initial Refresh needs catalogd.
All six G0 smokes (D1 through D6) PASS end-to-end after every fix
round. Phase G0 substrate is complete: 5 binaries, 6 routes, 25 fixes
applied across 6 days from cross-lineage review.
G1+ next: gRPC adapters, Lance/HNSW vector indices, Go MCP SDK port,
distillation rebuild, observer + Langfuse integration.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
9e9e4c26a4 |
G0 D5: queryd DuckDB SELECT over Parquet via httpfs · 4 scrum fixes
Phase G0 Day 5 ships queryd: in-memory DuckDB with custom Connector
that runs INSTALL httpfs / LOAD httpfs / CREATE OR REPLACE SECRET
(TYPE S3) on every new connection, sourced from SecretsProvider +
shared.S3Config. SetMaxOpenConns(1) so registrar's CREATE VIEWs and
handler's SELECTs serialize through one connection (avoids cross-
connection MVCC visibility edge cases).
Registrar.Refresh reads catalogd /catalog/list, runs CREATE OR
REPLACE VIEW "name" AS SELECT * FROM read_parquet('s3://bucket/key')
per manifest, drops views for removed manifests, skips on unchanged
updated_at (the implicit etag). Drop pass runs BEFORE create pass so
a poison manifest can't block other manifest refreshes (post-scrum
C1 fix).
POST /sql with JSON body {"sql":"…"} returns
{"columns":[{"name":"id","type":"BIGINT"},…], "rows":[[…]],
"row_count":N}. []byte → string conversion so VARCHAR rows
JSON-encode as text. 30s default refresh ticker, configurable via
[queryd].refresh_every.
Cross-lineage scrum on shipped code:
- Opus 4.7 (opencode): 1 BLOCK + 4 WARN + 4 INFO
- Kimi K2-0905 (openrouter): 2 BLOCK + 2 WARN + 1 INFO
- Qwen3-coder (openrouter): 2 BLOCK + 1 WARN + 1 INFO
Fixed (4):
C1 (Opus + Kimi convergent): Refresh aborts on first per-view error
→ drop pass first, collect errors, errors.Join. Poison manifest
no longer blocks the rest of the catalog from re-syncing.
B-CTX (Opus BLOCK): bootstrap closure captured OpenDB's ctx →
cancelled-ctx silently fails every reconnect. context.Background()
inside closure; passed ctx only for initial Ping.
B-LEAK (Kimi BLOCK): firstLine(stmt) truncated CREATE SECRET to 80
chars but those 80 chars contained KEY_ID + SECRET prefix → log
aggregator captures credentials. Stable per-statement labels +
redactCreds() filter on wrapped DuckDB errors.
JSON-ERR (Opus WARN): swallowed json.Encode error → silent
truncated 200 on unsupported column types. slog.Warn the failure.
Dismissed (4 false positives):
Qwen BLOCK "bootstrap not transactional" — DuckDB DDL is auto-commit
Qwen BLOCK "MaxBytesReader after Decode" — false, applied before
Kimi BLOCK "concurrent Refresh + user SELECT deadlock" — not a
deadlock, just serialization, by design with 10s timeout retry
Kimi WARN "dropView leaves r.known inconsistent" — current code
returns before the delete; the entry persists for retry
Critical reviewer behavior: 1 convergent BLOCK between Opus + Kimi
on the per-view error blocking, plus two independent single-reviewer
BLOCKs (B-CTX, B-LEAK) that smoke could never have caught. The
B-LEAK fix uses defense-in-depth: never pass SQL into the error
path AND redact known cred values from DuckDB's own error message.
DuckDB cgo path: github.com/duckdb/duckdb-go/v2 v2.10502.0 (per
ADR-001 §1) on Go 1.25 + arrow-go. Smoke 6/6 PASS after every
fix round.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
||
|
rootandClaude Opus 4.7
|
4205ecd0f0 |
Pre-D5: extract CatalogClient to internal/catalogclient/ + add List
queryd (D5) needs the same HTTP client to catalogd that ingestd uses, but the client lived in internal/ingestd — having queryd import from ingestd would invert the data-flow direction (ingestd is upstream of queryd; the package dep should not point back). Extract to a shared internal/catalogclient/ package now, before D5 forces it under implementation pressure. Adds the List(ctx) method queryd will need for view registration. Unit tests cover Register success/conflict and List success/error paths against an httptest.Server fake. ingestd's import flips from internal/ingestd → internal/catalogclient; the wire format and behavior are unchanged. All four smokes (D1/D2/D3/ D4) PASS unchanged. DuckDB cgo path re-verified with the official github.com/duckdb/duckdb-go/v2 (per ADR-001) on Go 1.25 + arrow-go. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |