lenny
AGENTS.md1 starsChanged 2 months agoMITView on GitHub
This repo vendors a portable skill stack under skills/. Any coding agent reading this file (Codex, zcode, Claude Code, etc.) should treat those skills as executable process documents and invoke them at the moments named below. Read a skill file in full before applying it. Skills compose: ship-conductor is the score and invokes the others at its phase gates. The ship-conductor skill reads this section: - Councils: for documentation, onboarding, or installation changes, convene stranger-installability / reference-integrity / safety-rails seats.…
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 31,700 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 repository20 other files it gives its agents.
- bayesianSkill
- claudes-gambleSkill
- codebase-designSkill
- councilSkill
- elon-first-principlesSkill
- find-skillsSkill
- first-principlesSkill
- fleet-conductorSkill
- garciaSkill
- gbrain-captureSkill
- grill-meSkill
- karpathySkill
- kissingerSkill
- meta-conductorSkill
- outcome-lockSkill
- pre-kSkill
- release-managerSkill
- rumsfeldSkill
- setup-lennySkill
- ship-conductorSkill
The file
We split this AGENTS.md into the rules it gives an agent
5 rules. About 2 are common, 3 are distinctive, and 0 are very specific to this project.
≈2
Common · estimate
Many files like this have a rule of this kind.
≈3
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
- could be automatic: a setting could enforce it
- Distinctive: line 22
Read a skill file in full before applying it.
- Common: line 36
Domain rails: never publish a repository or package, expose credentials, include private/internal material, or perform another …
- Distinctive: line 42
Commit surgically: stage only files you created/modified.
- Distinctive: line 44
Never merge, force-push, or rewrite shared history.
- Common: line 45
Never print, commit, or echo credentials.
≈ “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
AGENTS.md - Lenny project wiring
This repo vendors a portable skill stack under skills/. Any coding agent
reading this file (Codex, zcode, Claude Code, etc.) should treat those skills as
executable process documents and invoke them at the moments named below.
Skills
| Skill | File | Invoke when |
| setup-lenny | skills/setup-lenny/SKILL.md | user says “set up Lenny”; safely detects this repository and writes .lenny/profile.md |
| ship-conductor | skills/ship-conductor/SKILL.md | user says "conduct this plan"; drives plan to a merge-ready branch and STOPS before PR or merge |
| outcome-lock | skills/outcome-lock/SKILL.md | before a conducted run; freezes the user-visible outcome and completion criteria |
| grill-me | skills/grill-me/SKILL.md | before building: interrogate the plan one question at a time until shared understanding |
| first-principles | skills/first-principles/SKILL.md | always: no claim about reality without a verification path |
| elon-first-principles | skills/elon-first-principles/SKILL.md | scoping: delete relentlessly, park with concrete triggers |
| garcia | skills/garcia/SKILL.md | explicit "garcia" invocation: full-autonomy finish mode with hard rails |
| claudes-gamble | skills/claudes-gamble/SKILL.md | genuine design forks: adversarial multi-perspective panel |
| karpathy | skills/karpathy/SKILL.md | verification: run/measure it instead of reasoning about it |
| release-manager | skills/release-manager/SKILL.md | ship cutoff: freeze line, cut list with add-back triggers |
Read a skill file in full before applying it. Skills compose: ship-conductor is the score and invokes the others at its phase gates.
Conductor
The ship-conductor skill reads this section:
- Councils: for documentation, onboarding, or installation changes, convene
stranger-installability / reference-integrity / safety-rails seats. For the
evidence validator, convene correctness / failure-modes / portability seats.
Otherwise use the default Software Implementation Council at
skills/council/references/software-implementation-council.md. - Verification build:
npm run check. - Merge policy: branches and PRs only; the human opens the PR and merges.
- Domain rails: never publish a repository or package, expose credentials, include private/internal material, or perform another outward-facing action without explicit human approval.
Conventions for agents in this repo
- Commit surgically: stage only files you created/modified; verify staging before every commit.
- Never merge, force-push, or rewrite shared history; the human does that.
- Never print, commit, or echo credentials. Env references only.
- Long-running work keeps
CONDUCTOR-RUN.mdupdated on disk (see the ship-conductor skill) so any session can resume after context loss. lennytools/lennyis the canonical public core. Person- and company-specific councils or skills are optional private overlays, never public defaults.
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.