Skill Detail

Maintain reusable project knowledge with Project Cairn

Use Project Cairn to initialize, maintain, audit, and graduate project knowledge so coding-agent work leaves reusable decisions and lessons behind.

Templates & WorkflowsMulti-Framework
Templates & Workflows Multi-Framework Security Reviewed
⭐ 235 GitHub stars
COPY SKILL INSTRUCTIONS (OPTIONAL)
npx skills add agentskillexchange/skills --skill maintain-reusable-project-knowledge-with-project-cairn Copy
Uses the third-party skills CLI, not an ASE-owned installer. Check your agent’s compatibility. This copies instructions; complete the upstream tool setup below separately.
At a glance
Tools required
Skill-compatible agent such as Claude Code, Codex, WorkBuddy, or another runtime that can load a SKILL.md directory; git; bash for scripts/*.sh; Python 3 for scripts/*.py; optional PyYAML for notion-graduate-batch.py; optional Obsidian, Notion, or Lark / Feishu knowledge base for graduation workflows
Install & setup
Clone the repository into the skill directory for the target agent so SKILL.md sits at the root of project-cairn. For Claude Code, use git clone https://github.com/iBlinkQ/project-cairn.git ~/.claude/skills/project-cairn. For Codex, use git clone https://github.com/iBlinkQ/project-cairn.git ~/.agents/skills/project-cairn. In a project, ask the agent to initialize Project Cairn, then use the documented init, maintenance, audit, consume, graduation, upgrade, and branch-closure references as needed.
Author
iBlinkQ
Publisher
Individual
Last updated
Oct 5, 2026
Quick brief

Use Project Cairn when an operator wants an agent to run project-knowledge maintenance during day-to-day work instead of relying on chat history or informal notes. The agent can manage initialization or retrofit work for AGENTS.md, .cairn/config.yaml, and cairn/ notes; record meaningful progress in LOG.md and ROADMAP.md; review topic notes for current conclusions; audit stale or missing records; search and cite external knowledge only when it shapes the work; extract reusable lessons; and generate human-approved graduation candidates for Obsidian, Notion, or Lark / Feishu knowledge bases. Invoke this during agent-assisted project setup, after meaningful implementation or research turns, before closing exploration branches, or when a verified lesson should become reusable outside the current repository. The boundary is explicit: this is not a background memory product, generic note-taking app, CLI listing, or normal documentation framework. It is a skill-loaded operator workflow with documented triggers, templates, references, provider adapters, reviewable outputs, and human confirmation before long-term knowledge-base writes.