golems
AGENTS.md0 starsChanged 15 days agoApache-2.0View on GitHub
Bun workspace of reusable agent packages, CLIs, and evaluated workflow skills. The current workspace inventory and package purposes live in README.md. Before changing a package, read its CLAUDE.md if it has one. CI runs on PRs into any base branch, so a stacked PR gets the full suite (semantic mutants: input-changing PRs + nightly). Agents declaring isolation: worktree in .claude/agents/*.md get their own git worktree, for parallel work without file conflicts. See packages/claude/SOUL.md for bot persona. Durable content goes in…
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 31,774 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository64 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- verify-lead-waitSkill
- verify-merge-receiptsSkill
- branded-docSkill
- explanatory-docSkill
- 1passwordSkill
- agada-benchSkill
- agent-browserSkill
- agent-routingSkill
- architectural-conformance-auditSkill
- audio-dashboardSkill
- brain-store-fallbackSkill
- claude-desktop-researchSkill
- cmux-agentsSkill
- cmuxSkill
- coachSkill
- coderabbitSkill
- codex-workflowsSkill
- collab-monitorSkill
- content-demo-creationSkill
- convention-auditSkill
- cursor-workflowsSkill
- cyberSkill
- deploy-verifySkill
- drive-filingSkill
- eas-prebuild-checkSkill
- ecosystem-healthSkill
- eli5Skill
- false-green-gateSkill
- figma-loopSkill
- fleet-wrap-gateSkill
- fleet-wrapSkill
- frustration-captureSkill
- gemini-researchSkill
- git-guardianSkill
- github-researchSkill
- golem-installSkill
- grill-meSkill
- heredoc-guardSkill
- html-dashboardSkill
- human-confirm-gateSkill
- i-have-adhdSkill
- install-runbook-linterSkill
- large-planSkill
- m1-window-gateSkill
- model-pin-gateSkill
- never-fabricateSkill
- orcSkill
- plan-councilSkill
- pr-loopSkill
- qa-verdict-gateSkill
- qa-videoSkill
- repogolemSkill
- research-lifecycleSkill
- reviewer-order-gateSkill
- session-handoffSkill
- skill-creatorSkill
- stitch-designSkill
- t3-visualsSkill
- tmp-blockSkill
- The first 60 shown.
The file
We split this AGENTS.md into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
Before changing a package, read its CLAUDE.md if it has one.
≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- IDENTITY: golems monorepo, EtanHey, AI agent packages and skills -->Golems Monorepo
Bun workspace of reusable agent packages, CLIs, and evaluated workflow skills.
Packages
The current workspace inventory and package purposes live in README.md.
Before changing a package, read its CLAUDE.md if it has one.
Key Relationships
- ClaudeGolem preserves persona guidance and Claude CLI status workflows
- CoachGolem reads getStatus() from Teller (read-only)
- Services (briefing) imports from Coach for daily plan generation
- Cloud Worker remains a local/successor-host runnable scheduler; Railway service was deleted on 2026-07-05
- Most packages depend on Shared for Supabase, LLM, state; golem-skills, golems-tui, green-invoice-mcp and mock-mcp do not
Development
bun install # Install all workspace deps
bun run test # Package suite (bun test ./packages), as CI runs itCI runs on PRs into any base branch, so a stacked PR gets the full suite (semantic mutants: input-changing PRs + nightly).
Worktree-Isolated Agents
Agents declaring isolation: worktree in .claude/agents/*.md get their own git
worktree, for parallel work without file conflicts.
Communication Style
- Formality: 2/10 — Very casual
- Length: Brief, direct
- Tone: Friendly, sometimes playful
See packages/claude/SOUL.md for bot persona.
Never write durable content to /tmp
Durable content goes in the repo or its docs.local/ — never /tmp,
/private/tmp, /var/folders, or $TMPDIR, which are wiped on reboot and
invisible to the rest of the fleet. Worktrees go in <repo>/.worktrees/<name>.
Genuinely ephemeral? WEAVE_ALLOW_TMP=1 <command> — allowed and logged.
Claude and Cursor panes are held to this by the tmp-block PreToolUse hook;
Codex enforcement requires hooks-live install and owner trust review in plain codex;
a seat counts as guarded only after native denial is verified. The installer's --status reports wiring, trust unverified. Contract and limits:
skills/golem-powers/tmp-block/SKILL.md.
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.

