This release fixes issues for SREs watching stability and regressions.
Published 22d
AI Agents & Assistants
✓ No known CVEs patched
✓ No known CVEs patched in this version
Topics
agentic-ai
agentic-framework
agentic-workflow
agents
ai-agents
ai-assistant
+14 more
ai-coding
ai-skills
autonomous-agents
claude-code
codex
harness
mcp-server
multi-agent
multi-agent-systems
npm
skills
swarm
swarm-intelligence
typescript
Summary
AI summaryFixed two regressions: restored correct behavior for memory search and made npx ruflo init fail loudly if the optional memory dependency is missing.
Full changelog
Two external-reported regression fixes.
Fixed
- #2558 —
memory searchbroken (keyword recall random, HNSW index 0 vectors). Two defects: the Phase-2 BM25 fusion silently dropped the keyword-coverage floor (IDF collapses on small corpora → docs containing the query word scored below threshold), and WAL-mode writes were never checkpointed so WAL-blind readers saw 0 vectors. Fix: coverage floor + recall-friendly fusion +wal_checkpointon store + accuratetotal_vectors. Search now recalls confirmed-stored entries (targeted, not a dump). 169/169 memory suite. - #2545 —
npx ruflo initleft self-learning a silent no-op.@claude-flow/memory(optionalDependency) lands in the npx cache, off the project's node_modules walk-up, so the auto-memory hook never resolved it. Fix: init records the resolved path in a.claude-flow/memory-package.jsonsidecar; hook reads it first and now fails LOUD when memory is genuinely absent;doctor --fixrepairs it.
PRs: #2567, #2564
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
Related context
Related tools
Beta — feedback welcome: [email protected]