yourconscienceyourconscience/tackroom/AGENTS.md

tackroom

AGENTS.md8 starsChanged 35 days agoMITView on GitHub

This repository is the public tackroom CLI distribution. It must not contain a maintainer's personal skills, hooks, MCP servers, secrets, memory data, or machine-specific configuration. The CLI manages a separate user-owned configuration root: Never restore project-directory config discovery or make the public checkout the user's canonical configuration. The distribution intentionally contains: New starter content requires an explicit product decision and an inventory test update. Keep README.md, skills/tackroom/SKILL.md, CLI help, the public template, and release-site copy consistent. Examples must use conventional…

Installs packages

1 other project uses a version of this.

Adopted by

Most recently taken up by

yourconscience/dotagentsword for word, first seen Oct 8, 2026

Other versions1 project uses a version of this.
Word for word · 1
yourconscience/dotagents★ 8100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
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 repository3 other files it gives its agents.
- memory/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- tackroomSkill

The file

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

9 rules. About 1 is common, 6 are distinctive, and 2 are very specific to this project.

≈1
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
2
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:
Never restore project-directory config discovery or make the public checkout the user's canonical configuration.
line 11
Very specific:
Use focused package tests while iterating.
line 43
Distinctive:
Use Python 3's standard library for the dependency-free basic memory tier.
line 44
Distinctive:
Keep tests isolated with temporary homes and paths.
line 45
Distinctive:
Prefer surgical changes and existing adapters over parallel implementations.
line 46
Distinctive:
Do not add new unit tests.
line 47
Very specific:
Keep README.md, skills/tackroom/SKILL.md, CLI help, the public template, and release-site copy consistent.
line 52
Distinctive:
Use short, single-line commit messages without bot attribution trailers.
line 57
Common:
Do not commit secrets, local overlays, external caches, generated memory, or native harness state.
line 58

≈ “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

Project scope

This repository is the public tackroom CLI distribution. It must not contain a maintainer's personal skills, hooks, MCP servers, secrets, memory data, or machine-specific configuration.

The CLI manages a separate user-owned configuration root:

--config /path/to/tackroom.yaml, when supplied.
$TACKROOM_HOME/tackroom.yaml, when set.
~/.agents/tackroom.yaml otherwise.

Never restore project-directory config discovery or make the public checkout the user's canonical configuration.

Architecture invariants

tackroom syncs exactly five surfaces: skills, MCP servers, hooks, agent roles, and plugins.
Plugins follow the agent-plugins-spec v1.0.0 format. For harnesses with native plugin systems (Codex), tackroom projects spec-format plugins into the native format. Claude Code plugins stay in Claude's native layout: the config root's .claude-plugin/marketplace.json is registered as a directory marketplace and its relative-path plugins are enabled in ~/.claude/settings.json.
setup owns first-run scaffolding, harness detection, optional copy/convert import, memory-tier selection, and the first sync. There is no separate init command.
Native import is copy-only. Never move or delete the source content.
Existing unrelated native harness configuration must remain untouched.
External Git skills are commit-pinned in tackroom.lock; materialized files remain owned by that lock entry.
Harness adapters must represent verified native capabilities. Do not invent unsupported MCP, hook, role, or skill surfaces.

Public starter inventory

The distribution intentionally contains:

skills/tackroom/
starter.gitignore (written as the config root's .gitignore)
agents/architect.md
agents/builder.md
agents/general.md
agents/researcher.md
agents/reviewer.md
agents/tester.md
reusable memory infrastructure under memory/

New starter content requires an explicit product decision and an inventory test update.

Development

Go version and module layout are declared by go.mod. The memory tools under memory/tools/ are packages of that root module; setup scaffolds a standalone go.mod beside each copied tool so a user-owned config root can build them without the source checkout.
Install the development CLI with go install ./cmd/tackroom.
Use focused package tests while iterating; run go test ./... before submission.
Use Python 3's standard library for the dependency-free basic memory tier.
Keep tests isolated with temporary homes and paths. Tests must never mutate live harness configuration.
Prefer surgical changes and existing adapters over parallel implementations.
Do not add new unit tests. When a change breaks an existing unit test, delete that test instead of updating it.
TODO: replace the unit tests with end-to-end checks that tackroom setup and tackroom sync install correctly for selected harness setups in a temporary home.

Documentation

Keep README.md, skills/tackroom/SKILL.md, CLI help, the public template, and release-site copy consistent. Examples must use conventional paths or environment variables, not one maintainer's filesystem.

Git and review

Preserve unrelated worktree changes.
Use short, single-line commit messages without bot attribution trailers.
Do not commit secrets, local overlays, external caches, generated memory, or native harness state.
Pull requests must pass the full repository suite before merge.

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