Skip to content

This release adds 3 notable features for engineering teams evaluating rollout.

Published 16d MCP Developer Tools
✓ No known CVEs patched
Read the diff → Tool health → What is this tool? →

✓ No known CVEs patched in this version

Topics

agent-memory ai-agents claude cline codex coding-agents
+10 more
cursor developer-tools llm local-first mcp mcp-server model-context-protocol persistent-memory python sqlite

Summary

AI summary

Added a fixed 6-case v0.20 clean‑room proof pack, an executable runner script, and generated snapshots.

Changes in this release

Feature Low

Added a fixed 6-case v0.20 clean-room proof pack.

Added a fixed 6-case v0.20 clean-room proof pack.

Source: llm_adapter@2026-07-15

Confidence: high

Feature Low

Added an executable runner: `python scripts/run_v020_clean_room_proof.py`.

Added an executable runner: `python scripts/run_v020_clean_room_proof.py`.

Source: llm_adapter@2026-07-15

Confidence: high

Feature Low

Added generated snapshots: `benchmark/latest-v0.20-clean-room-proof-report.json` and `benchmark/latest-v0.20-clean-room-proof-transcript.md`.

Added generated snapshots: `benchmark/latest-v0.20-clean-room-proof-report.json` and `benchmark/latest-v0.20-clean-room-proof-transcript.md`.

Source: llm_adapter@2026-07-15

Confidence: high

Feature Low

Added release-contract facts for the v0.20 denominator and pass rates.

Added release-contract facts for the v0.20 denominator and pass rates.

Source: llm_adapter@2026-07-15

Confidence: high

Bugfix Medium

Public MCP surface unchanged at 10 tools; validation passed with 336 pytest tests.

Public MCP surface unchanged at 10 tools; validation passed with 336 pytest tests.

Source: llm_adapter@2026-07-15

Confidence: high

Full changelog

v0.20.0 - Clean-Room Adoption Proof

Agent Memory Bridge 0.20.0 is a clean-room proof release over the v0.19
proof-breadth work.

0.20.0 = clean-room adoption before product expansion.

What Changed

  • Added a fixed 6-case v0.20 clean-room proof pack:
    • local entrypoint/import sanity
    • real stdio MCP tool-surface check
    • tokened store -> recall over MCP stdio
    • first-run CLI rendering
    • Task Brief CLI rendering
    • isolated temp-store write-scope guard
  • Added an executable runner:
    • python scripts/run_v020_clean_room_proof.py
  • Added generated snapshots:
    • benchmark/latest-v0.20-clean-room-proof-report.json
    • benchmark/latest-v0.20-clean-room-proof-transcript.md
  • Added release-contract facts for the v0.20 denominator and pass rates.
  • Kept the public MCP surface unchanged at 10 tools.

Validation

  • pytest: 336 passed
  • release contract: passed
  • public MCP surface: unchanged at 10 tools
  • v0.20 clean-room proof pack: 6 / 6 cases passed

Boundaries

The v0.20 clean-room proof is local reproducible evidence. It is not external
vendor certification.

This release is still not:

  • a new MCP tool
  • an AMH runtime dependency
  • a client config writer
  • a plugin package for every client
  • a scheduler, watcher, worker, daemon, or runtime loop
  • an automatic promotion, deletion, merge, approval, or writeback path

Public MCP Surface

The public MCP surface remains unchanged:

  • store
  • recall
  • browse
  • stats
  • forget
  • promote
  • export
  • claim_signal
  • extend_signal_lease
  • ack_signal

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

Share this release

Track zzhang82/Agent-Memory-Bridge

Get notified when new releases ship.

Sign up free

About zzhang82/Agent-Memory-Bridge

MCP-native, local-first memory for coding agents that turns coding sessions into reusable engineering memory: decisions, gotchas, and domain knowledge.

All releases →

Related context

Earlier breaking changes

  • v0.21.0 Procedures with declared domains are rejected for mismatched task domains; legacy unscoped procedures now emit a warning instead of auto‑assigning policy.

Beta — feedback welcome: [email protected]