This is the operational contract for how Claude should interact with a personal knowledge base. It enforces a brain-first lookup protocol: check what you already know before hitting external APIs, write back everything you learn with inline source citations, and maintain bidirectional links between entities. The key insight here is treating the knowledge base as an ambient context layer rather than a separate retrieval step. Every conversation becomes a read-enrich-write loop. User statements are treated as the highest-authority data source. The iron law is back-linking: every mention of an entity must create a link from that entity's page. Useful if you're building a persistent context system and want consistent conventions around citation, enrichment timing, and graph maintenance.

skills.sh installs
419
GitHub stars
32.3K
last updated
Aug 21, 2026

Install

$ npx -y skills add garrytan/gbrain --skill brain-ops --agent claude-code

Installs into .claude/skills of the current project.

Files

Brain Operations — The Ambient Context Layer

Recall relevant context before responding. Save explicit requests with provenance; automatic capture is off until the user opts in. Reading this skill does not enable capture, delegation, or paid enrichment. A chat-only instruction suppresses writes for that turn, including when standing capture is enabled.

Convention: See skills/conventions/brain-first.md for the lookup chain and copyable retrieval filters. Convention: See skills/conventions/quality.md for citation and back-link rules.

Contract

This skill guarantees:

  • Brain is checked BEFORE any external API call (brain-first lookup)
  • Explicit save requests and opted-in inbound signals trigger the READ → WRITE loop; enrichment requires its separately configured authority
  • Every outbound response checks brain for relevant context
  • Source attribution on every fact written (inline [Source: ...] citations)
  • User's direct statements are highest-authority data
  • Back-links maintained on every brain write (Iron Law)

Iron Law: every mention of a person or company with a brain page gets a back-link FROM that entity's page TO the page mentioning them.

Sponsors

Details

Repository
github.com/garrytan/gbrain
Category
Design & UI/UX
First seen
Jun 3, 2026

More from garrytan/gbrain

All 2
1aj-geddes avatarmulti-cloud-strategyaj-geddes/useful-ai-promptsDesign and implement multi-cloud strategies spanning AWS, Azure, and GCP with vendor lock-in avoidance, hybrid deployments, and federation.418skills.sh installs
2davila7 avatarworkflow-automationdavila7/claude-code-templatesWorkflow automation is the infrastructure that makes AI agents reliable. Without durable execution, a network hiccup during a 10-step payment flow means lost mo416skills.sh installs
3onewave-ai avataranimateonewave-ai/claude-skillsGenerate animated videos and motion graphics from natural language descriptions. Creates a standalone Vite + React project with Framer Motion scenes that auto-p413skills.sh installs
4mims-harvard avatartooluniverse-spatial-transcriptomicsmims-harvard/tooluniverseSpatial transcriptomics analysis — Visium, MERFISH, seqFISH, Slide-seq. Maps gene expression to tissue architecture, identifies spatially variable genes (SVGs),411skills.sh installs
5mims-harvard avatartooluniverse-rnaseq-deseq2mims-harvard/tooluniverseRNA-seq differential expression analysis with DESeq2, edgeR, and limma-voom — DEG lists, fold changes, dispersion estimation, design formulas including covariat409skills.sh installs
6hoodini avatarmermaid-diagramshoodini/ai-agents-skillsCreate diagrams and visualizations using Mermaid syntax. Use when generating flowcharts, sequence diagrams, class diagrams, entity-relationship diagrams, Gantt 408skills.sh installs