ai-memory (0.6.0-1) noble; urgency=medium

  * Phase 1 complete: Tasks 1.1 schema metadata, 1.2 agent identity (NHI),
    1.3 agent registration + _agents namespace, 1.5 scope-based visibility,
    1.6 N-level namespace standards, 1.7 vertical memory promotion,
    1.8 horizontal forget, 1.9 per-namespace governance + consensus,
    1.10 pending-action approvals, 1.11 budget-aware recall,
    1.12 hierarchy-aware recall + scope_idx + proximity boost
  * Peer-mesh sync-daemon with vector clocks + timestamp-aware merge
  * Native TLS + mTLS fingerprint allowlist (Layer 1 + 2)
  * v0.6.0.0 sprint: time-decay half-life on recall scoring,
    contextual recall via context-token embedding fusion,
    post-store LLM autonomy hooks (auto_tag + detect_contradiction),
    TypeScript + Python SDK scaffolds, hardened systemd units
  * 23 MCP tools, HTTP API with 24+ endpoints, CLI with 26+ commands
  * MANDATORY disclosures: consensus governance breaking change (#234),
    sync-endpoint TLS requirement (#231), mTLS trust boundary (#239).
    Read CHANGELOG.md before deploying.

 -- Justin Jessup <Justin@alpha-one.mobi>  Sun, 19 Apr 2026 00:00:00 +0000

ai-memory (0.5.1-1) noble; urgency=medium

  * Docker image on GHCR, auto-built on tag push
  * Official MCP Registry published (io.github.alphaonedev/ai-memory)
  * Dockerfile modernized: Rust 1.86, build-essential, MCP registry label
  * ARM64 Linux packages fixed (nfpm arch mapping)
  * Documentation red team audit: 6 critical fixes across all docs
  * OpenClaw added as 9th supported AI platform
  * New files: CONTRIBUTING.md, CHANGELOG.md, CODE_OF_CONDUCT.md
  * GitHub Pages: Open Graph tags, scope tables, professionalized hero

 -- Justin Jessup <Justin@alpha-one.mobi>  Tue, 08 Apr 2026 12:00:00 +0000

ai-memory (0.5.0-1) noble; urgency=medium

  * Initial release
  * MCP server with 17 tools for AI-native memory management
  * HTTP API with 20 endpoints
  * CLI with 25 commands
  * 4 feature tiers: keyword, semantic, smart, autonomous
  * Hybrid recall with FTS5 + semantic embeddings
  * TOON compact format (79% smaller than JSON)
  * 97.8% R@5 on ICLR 2025 LongMemEval benchmark

 -- Justin Jessup <Justin@alpha-one.mobi>  Tue, 08 Apr 2026 00:00:00 +0000
