memory-mcp
yuvalsuede/memory-mcp
An intelligent, two-tier persistent memory and automated git snapshot system for Claude Code that captures project decisions and architecture in the background.
98 12MITUpdated 2026-01-29
71
GoodOverview
memory-mcp acts as an automated memory layer for AI coding sessions in Claude Code. Using local hooks and a lightweight LLM (Haiku), it silently extracts architectural choices, patterns, and progress updates from conversation transcripts. It maintains an instant-load CLAUDE.md file for startup context and a comprehensive searchable JSON store for deep recall via MCP tools, alongside automatic project versioning via hidden git snapshots.
Capabilities
- ▸Automatic silent memory extraction via session hooks
- ▸Two-tier memory storage (CLAUDE.md and .memory/state.json)
- ▸LLM-powered memory consolidation, deduplication, and confidence decay
- ▸MCP server tools for deep search, Q&A, and memory management
- ▸Automated git snapshotting and rollback on a separate branch
- ▸Visual terminal and HTML context dashboards
Best for
Retaining project architecture and technical decisions across multiple AI coding sessions, Instant project recovery and rollback using automated session-linked git snapshots, Searching deep project context mid-conversation using MCP tools, Preventing context window loss and the need to repeatedly re-explain project layouts
Works with
Claude Code
Evoa Score breakdown= Σ (score × weight)
71Task usefulness20%100 → +20.0
実タスクにどれだけ役立つか（機能の豊富さ・用途の明確さ）。 — AIによるcapabilities/use-cases解析
Code quality15%90 → +13.5
実装・指示の品質。 — AIによるSKILL.md/README解析
Maintenance15%50 → +7.5
リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ
Documentation12%37 → +4.4
ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量
Security15%80 → +12.0
危険・不審な挙動が無いか。 — AIによるセキュリティレビュー
Originality10%85 → +8.5
ありふれたラッパーではない独自性。 — AIによる独自性判定
Popularity8%52 → +4.2
コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）
Compatibility5%25 → +1.3
対応AIエージェントの広さ。 — AIによる対応エージェント判定
Weighted total71.3 / 100
ライセンス不明/制限あり（Red）のSkillは総合スコアに0.85倍の補正を適用します。 ランキングはこのScoreのみで決まり、広告で変わりません。 算出方法の詳細 →
Security considerations
Requires access to an Anthropic API key to process transcripts externally via Haiku, and commits project state to a hidden local git branch.
Categories
Summary and analysis are original content generated by AI Skills Rank. The skill's source text is not reproduced here — view it on the linked repository.