Initial commit: Agent Governance System Phase 8

Phase 8 Production Hardening with complete governance infrastructure:

- Vault integration with tiered policies (T0-T4)
- DragonflyDB state management
- SQLite audit ledger
- Pipeline DSL and templates
- Promotion/revocation engine
- Checkpoint system for session persistence
- Health manager and circuit breaker for fault tolerance
- GitHub/Slack integrations
- Architectural test pipeline with bug watcher, suggestion engine, council review
- Multi-agent chaos testing framework

Test Results:
- Governance tests: 68/68 passing
- E2E workflow: 16/16 passing
- Phase 2 Vault: 14/14 passing
- Integration tests: 27/27 passing

Coverage: 57.6% average across 12 phases

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

bin/memory

@@ -0,0 +1,106 @@
+#!/bin/bash
+#
+# External Memory Layer CLI for Claude Code
+# ==========================================
+# Token-efficient storage and retrieval of large outputs, transcripts, and context.
+#
+# Commands:
+#   log <content>              Store content in memory (auto-chunks large content)
+#   log --file <path>          Store file contents
+#   log --stdin                Read from stdin
+#   fetch <id>                 Retrieve memory entry
+#   fetch <id> --summary-only  Get just the summary
+#   fetch <id> --chunk N       Get specific chunk
+#   list [--type TYPE]         List memory entries
+#   search <query>             Search memory by content/tags
+#   summarize <id>             Generate/show summary
+#   refs --checkpoint <id>     Get memory refs for checkpoint
+#   prune                      Clean old entries
+#   stats                      Show memory statistics
+#
+# Examples:
+#   echo "Large output..." | memory log --stdin --tag "test"
+#   memory fetch mem-20260123-123456-abcd --summary-only
+#   memory list --type output --limit 10
+#   memory search "error"
+#
+# Token Guidelines:
+#   - Content < 500 tokens: stored inline
+#   - Content 500-4000 tokens: stored in file with summary
+#   - Content > 4000 tokens: auto-chunked with parent summary
+#
+# Integration:
+#   - Link to checkpoints: --checkpoint <id>
+#   - Link to directories: --directory <path>
+#   - Summaries auto-generated for efficient retrieval
+#
+# Documentation: /opt/agent-governance/docs/MEMORY_LAYER.md
+
+set -euo pipefail
+
+MEMORY_SCRIPT="/opt/agent-governance/memory/memory.py"
+
+# Show help if no args or --help
+if [[ $# -eq 0 ]] || [[ "${1:-}" == "--help" ]] || [[ "${1:-}" == "-h" ]]; then
+    cat << 'EOF'
+External Memory Layer
+=====================
+
+Commands:
+  log <content>              Store content in memory
+  log --file <path>          Store file contents
+  log --stdin                Read from stdin
+  fetch <id>                 Retrieve memory entry
+  fetch <id> -s              Summary only
+  fetch <id> --chunk N       Get specific chunk
+  list [--type TYPE]         List entries
+  search <query>             Search memory
+  summarize <id>             Generate summary
+  refs --checkpoint <id>     Memory refs for checkpoint
+  prune                      Clean old entries
+  stats                      Show statistics
+
+Options for 'log':
+  --file, -f <path>          Read from file
+  --stdin                    Read from stdin
+  --type, -t <type>          Type: transcript, output, context
+  --tag <tag>                Add tag (repeatable)
+  --checkpoint <id>          Link to checkpoint
+  --directory, -d <path>     Link to directory
+  --no-chunk                 Don't auto-chunk large content
+  --json                     Output JSON
+
+Token Thresholds:
+  < 500 tokens    → stored inline
+  500-4000 tokens → stored in file + summary
+  > 4000 tokens   → auto-chunked + summary
+
+Examples:
+  # Store large test output
+  pytest tests/ 2>&1 | memory log --stdin --tag "pytest" --tag "tests"
+
+  # Fetch just the summary
+  memory fetch mem-20260123-123456-abcd -s
+
+  # Get chunk 2 of a large entry
+  memory fetch mem-20260123-123456-abcd --chunk 2
+
+  # List recent outputs
+  memory list --type output --limit 5
+
+  # Search for errors
+  memory search "failed" --limit 10
+
+Documentation: /opt/agent-governance/docs/MEMORY_LAYER.md
+EOF
+    exit 0
+fi
+
+# Verify memory script exists
+if [[ ! -f "${MEMORY_SCRIPT}" ]]; then
+    echo "Error: memory.py not found at ${MEMORY_SCRIPT}"
+    exit 1
+fi
+
+# Execute memory command
+exec python3 "${MEMORY_SCRIPT}" "$@"