vintavinta/hal-9000/AGENTS.md

hal-9000

AGENTS.md142 starsChanged 22 days agoMITView on GitHub

macOS dev environment automation: dotfiles, AI agent configs, skills, and dev stacks. Run make help to list targets and hal --help for the CLI. Use make targets instead of running the underlying commands directly. They chain the right tools with the right flags. Pre-resolved Context7 IDs for the find-docs skill. Pass them to ctx7 docs and skip ctx7 library:

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 repository14 other files it gives its agents.
- publish-pluginsSkill
- upgrade-playbooksSkill
- audit-agent-settingsSkill
- best-practicesSkill
- blindspotSkill
- commitSkill
- differenceSkill
- fuck-over-engineeringSkill
- prSkill
- refactor-agents-mdSkill
- refactor-memorySkill
- refactor-skillSkill
- simple-englishSkill
- write-like-meSkill

The file

We split this AGENTS.md into the rules it gives an agent

3 rules. About 0 are common, 0 are distinctive, and 3 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Very specific:
Run make help to list targets and hal --help for the CLI.
line 7
Very specific:
Use make targets instead of running the underlying commands directly.
line 9
Very specific:
All skill descriptions must start with Use when, Use before, or Use after, except skills with disable-model-invocation: true, whose …
line 16

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

AGENTS.md

macOS dev environment automation: dotfiles, AI agent configs, skills, and dev stacks.

Commands

Run make help to list targets and hal --help for the CLI.

Use make targets instead of running the underlying commands directly. They chain the right tools with the right flags.

Gotchas

Edit under dotfiles/, never under ~/: hal sync symlinks every links entry in dotfiles/hal_dotfiles.json into ~/, so an edit to a linked path is live immediately. A new file outside a linked directory needs its own entry plus hal sync before anything references it.
Mirror dotfiles/.claude/settings.json edits into dotfiles/.claude/settings.sandbox.json: the sandbox copy is the settings for Claude Code under blackwall and drifts otherwise. Skip what it leaves out on purpose: hooks, the ask list, host-path deny rules, Ollama-backed plugins and env vars, and its own theme and statusline paths.
Edits under skills/ and plugins/ are not live until published: Claude Code loads them from the hal-9000 marketplace on GitHub (see dotfiles/.claude/settings.json), and other coding agents install skills/ via npx skills add vinta/hal-9000. Publish with a version bump (the publish-plugins skill) for a change to reach either.
All skill descriptions must start with Use when, Use before, or Use after, except skills with disable-model-invocation: true, whose description is a human-facing summary. A project-level skill's description may have a (project) prefix.
For generated artifacts such as zsh completion, regenerate them with the repo command instead of editing them by hand (e.g. make hal-completion after modifying bin/hal.py).

External Tool Documentation

Pre-resolved Context7 IDs for the find-docs skill. Pass them to ctx7 docs and skip ctx7 library:

Tool	`libraryId`
ansible	`/websites/ansible_projects_ansible`
ansible-lint	`/ansible/ansible-lint`
betterleaks	`/betterleaks/betterleaks`
fnm	`/schniz/fnm`
ghostty	`/websites/ghostty`
github-actions	`/websites/github_en_actions`
homebrew	`/homebrew/brew`
oh-my-zsh	`/ohmyzsh/ohmyzsh`
ollama	`/ollama/ollama`
pre-commit	`/pre-commit/pre-commit.com`
pytest	`/pytest-dev/pytest`
ruff	`/websites/astral_sh_ruff`
ty	`/websites/astral_sh_ty`
uv	`/websites/astral_sh_uv`
zsh	`/websites/zsh_sourceforge_io_doc_release`

Download now (ZIP)

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

What did you use it for? e.g. Reviewing pull requests in a Rails app

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.

Check your setup, free