gbrain
garrytan/gbrain
GBrain is a persistent memory and institutional knowledge engine for AI agents that provides synthesis, graph traversal, and gap analysis. Created by Garry Tan, it functions as a personal or company brain layer with privacy controls, vector search, and self-wiring knowledge graphs.
28.4k 4.2kMITUpdated 2026-08-13
83
GreatOverview
GBrain serves as an advanced second brain and context layer for AI coding agents and autonomous workflows. It aggregates information from diverse sources (meetings, emails, notes), extracts entity relationships without extra LLM overhead, and performs automated synthesis and gap analysis. It supports multiple backends including PGLite and Postgres with pgvector, multi-tenant workspace scoping, and continuous self-consolidation background jobs.
Capabilities
- ▸Synthesized multi-source question answering with explicit citations
- ▸Self-wiring knowledge graph extraction with typed edges
- ▸Vector search and semantic retrieval benchmarked for high precision
- ▸Multi-tenant company brain scoping and security boundaries
- ▸Automated trajectory tracking and scorecards for entities and metrics
Best for
Maintaining long-term persistent memory for AI coding agents, Serving as a shared company institutional memory with multi-user isolation, Tracking founder, company, and deal trajectories over time with metric-based scoring, Automating ingestion, enrichment, and consolidation of unstructured notes and logs
Works with
Claude CodeCodexCursor
Evoa Score breakdown= Σ (score × weight)
83Task usefulness20%88 → +17.6
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%100 → +15.0
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%42 → +5.0
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%100 → +8.0
コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）
Compatibility5%75 → +3.8
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 total83.4 / 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
GBrain distinguishes trusted local CLI callers from untrusted agent-facing remote callers via an explicit trust boundary. Sensitive operations like file uploads enforce strict filesystem confinement for remote callers, and protected phase handlers are restricted to local execution.
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.