agentleFS
Sign inStart free
juanitourquizajuanitourquiza/shipframe/AGENTS.md

shipframe

AGENTS.md4 starsChanged 19 days agoMITView on GitHub

ShipFrame is an AI coding workflow toolkit for teams that plan, prove, and ship software changes. Run before committing installer/workflow changes: --doctor --repo-only is the CI-safe gate. It must not depend on user auth, installed Claude/OpenCode/Codex binaries, or global HOME state. Before publishing, confirm upstream licensing for the Axis-Human base and keep THIRD_PARTY_NOTICES.md current.

The file

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

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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
  1. Very specific:

    Keep ShipFrame core generic and team-friendly.

    line 28
  2. Very specific:

    Put project-specific behavior in .shipframe/profile.md, shipframe.profile.md, or project-packs/.

    line 29
  3. Distinctive:

    Do not hardcode PULSAI or any client-specific release behavior into public core skills.

    line 30
  4. Distinctive:

    Keep documentation current in the same change: whenever installer behavior, user-facing commands, workflow routing, public skill behavior, …

    line 33
  5. Distinctive:

    Run before committing installer/workflow changes:

    line 37
  6. Very specific:

    Before publishing, confirm upstream licensing for the Axis-Human base and keep THIRD_PARTY_NOTICES.md current.

    line 52

≈ “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

ShipFrame Repository Instructions

ShipFrame is an AI coding workflow toolkit for teams that plan, prove, and ship software changes.

Stack

  • Markdown-first skills and agents.
  • Bash installer (install.sh).
  • Node.js snippets embedded inside install.sh for JSON editing and agent conversion.
  • Claude Code plugin metadata in .claude-plugin/.
  • Hermes Agent plugin packaging in plugin.yaml + __init__.py.
  • Codex workflow block in codex/dev-workflow.md.

Repo structure

  • skills/<name>/SKILL.md — installable skills. Keep this flat; the installer symlinks each direct child into host skill directories (~/.agents/skills, ~/.codex/skills, ~/.config/opencode/skills).
  • plugin.yaml + __init__.py — Hermes Agent plugin manifest and entrypoint at the repo root; __init__.py registers each skill under skills/ as a namespaced shipframe:<skill> Hermes skill. Dependency-free (stdlib only); keep it that way.
  • agents/*.md — Claude/OpenCode-oriented agents.
  • codex/dev-workflow.md — Codex routing workflow inlined into ~/.codex/AGENTS.md by the installer.
  • hooks/prompt-router-core.cjs — shared advisory bypass/suggest/route prompt classifier.
  • opencode/ — OpenCode v2 plugin adapter for ephemeral prompt-routing context.
  • templates/ — PR, issue, ClickUp, and wiki templates.
  • project-packs/ — optional project profile starters; do not hardcode these rules into generic core skills.
  • wiki/ and WIKI.md — local knowledge wiki.

Conventions

  • Keep ShipFrame core generic and team-friendly.
  • Put project-specific behavior in .shipframe/profile.md, shipframe.profile.md, or project-packs/.
  • Do not hardcode PULSAI or any client-specific release behavior into public core skills.
  • Preserve support for Claude Code, Codex CLI, and OpenCode when changing installer or workflow files.
  • For release-related behavior, require concrete deploy evidence before declaring work complete.
  • Keep documentation current in the same change: whenever installer behavior, user-facing commands, workflow routing, public skill behavior, release process, or project conventions change, update README and any affected repo docs before committing/releasing.

Verification

Run before committing installer/workflow changes:

bash -n install.sh
shellcheck install.sh
./install.sh --doctor --repo-only
./tests/test-install.sh
node tests/test-routing-parity.js # required when changing routing
python3 scripts/check-contracts.py
claude plugin validate .

--doctor --repo-only is the CI-safe gate. It must not depend on user auth, installed Claude/OpenCode/Codex binaries, or global HOME state.

Before publishing, confirm upstream licensing for the Axis-Human base and keep THIRD_PARTY_NOTICES.md current.

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