labmate
AGENTS.md25 starsChanged 2 months agoMITView on GitHub
Research harness for AI coding agents: 12 portable skills, five Claude named agents, shared references/scripts, and four default hook handlers across four lifecycle events. - Codex does not register agents/*.md. Agent-backed skills follow references/agent-routing.md: named agent, generic subagent, then main-thread fallback. - Deterministic project mutation belongs in scripts/; SKILL.md files remain thin discovery and judgment guides. - Claude explicit-invocation policy lives in skill frontmatter. Codex policy lives in skills/*/agents/openai.yaml. - Package and both plugin manifests share one version. A changed…
In this repository15 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- project-skillSkill
- project-skillSkill
- analyze-experimentSkill
- commit-changelogSkill
- init-projectSkill
- monitorSkill
- new-experimentSkill
- read-paperSkill
- survey-literatureSkill
- todoSkill
- update-docsSkill
- update-knowhowSkill
- update-project-skillSkill
- visualizeSkill
The file
Files
LabMate
Research harness for AI coding agents: 12 portable skills, five Claude named agents, shared references/scripts, and four default hook handlers across four lifecycle events.
Non-obvious boundaries
- Codex does not register
agents/*.md. Agent-backed skills followreferences/agent-routing.md: named agent, generic subagent, then main-thread fallback. - Deterministic project mutation belongs in
scripts/; SKILL.md files remain thin discovery and judgment guides. - Claude explicit-invocation policy lives in skill frontmatter. Codex policy
lives in
skills/*/agents/openai.yaml. - Package and both plugin manifests share one version. A changed plugin must receive a new version so installed caches cannot alias different source.
Verification
bash tests/test-hooks.sh
bash tests/test-platform-compat.sh
bash tests/test-context-contract.sh
bash tests/test-scripts.shPreserve unrelated worktree changes. Commit this nested repository before its parent gitlink.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.