agentleFS
Sign inStart free
opendatahub-ioopendatahub-io/docs-skills/AGENTS.md

docs-skills

AGENTS.md4 starsChanged 4 months agoApache-2.0View on GitHub

A Claude Code plugin providing documentation review, writing, and workflow tools. This file defines the shared project conventions for all AI coding agents. For Claude Code-specific instructions, see CLAUDE.md. The runtime working directory is the project root, not the skill directory. Bare relative paths like scripts/foo.py will fail. Always use a substitution variable: Use paths relative to the repository root (workspace): Skills (invoked via the Skill tool) use bare names: docs-workflow-requirements, jira-reader, learn-code. Qualified names (docs-skills:docs-workflow-requirements) also work. Use bare…

  • Installs packages

The file

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

5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    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
  • could be automatic: a setting could enforce it
  1. Distinctive:

    Use paths relative to the repository root (workspace):

    line 38
  2. Distinctive:

    Use kebab-case for skill and agent names

    line 52
  3. Very specific:

    Run make lint before committing (skillsaw + ruff + shellcheck)

    line 56
  4. Very specific:

    Use feat:, fix:, docs:, chore: commit prefixes

    line 58
  5. Very specific:

    When referencing Python in install steps, always use python3

    line 59

≈ “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

docs-skills

A Claude Code plugin providing documentation review, writing, and workflow tools. This file defines the shared project conventions for all AI coding agents. For Claude Code-specific instructions, see CLAUDE.md.

Repository structure

.claude-plugin/plugin.json   Plugin packaging metadata (name, version, description)
skills/<skill>/SKILL.md      Skill definitions with frontmatter
agents/<agent>.md            Subagent definitions with frontmatter
reference/                   Shared domain knowledge (frameworks, templates, guides)
hooks/hooks.json             Plugin-level Claude Code event hooks
eval/                        Evaluation test cases and harness config

Calling scripts from skills

The runtime working directory is the project root, not the skill directory. Bare relative paths like scripts/foo.py will fail. Always use a substitution variable:

Claude Code

  • ${CLAUDE_SKILL_DIR} — the directory containing the skill's SKILL.md. Use for scripts bundled with the same skill.
  • ${CLAUDE_PLUGIN_ROOT} — the plugin's installation directory (repo root). Use for cross-skill calls.
# Same-skill call (stdlib-only script)
python3 ${CLAUDE_SKILL_DIR}/scripts/detect_language.py --repo /path/to/repo

# Same-skill call (PEP 723 script with external deps)
uv run --script ${CLAUDE_SKILL_DIR}/scripts/jira_reader.py --issue PROJ-123

# Cross-skill call
python3 ${CLAUDE_PLUGIN_ROOT}/skills/learn-code/scripts/detect_language.py --repo /path

Cursor

Use paths relative to the repository root (workspace):

python3 skills/learn-code/scripts/detect_language.py --repo /path/to/repo

Skill and agent naming

Skills (invoked via the Skill tool) use bare names: docs-workflow-requirements, jira-reader, learn-code. Qualified names (docs-skills:docs-workflow-requirements) also work. Use bare names in workflow YAML step lists and skill-to-skill invocations.

Agents (invoked via the Agent tool's subagent_type) require fully-qualified names with the plugin prefix: docs-skills:technical-reviewer, docs-skills:docs-writer. Bare names like technical-reviewer will fail with "Agent type not found".

Contributing rules

  • Use kebab-case for skill and agent names
  • Bump version in .claude-plugin/plugin.json when making changes
  • New Python scripts with external dependencies must use PEP 723 inline metadata
  • New stdlib-only scripts use plain python3 invocation
  • Run make lint before committing (skillsaw + ruff + shellcheck)
  • Install test dependencies with pip install -r requirements.txt before running make test
  • Use feat:, fix:, docs:, chore: commit prefixes
  • When referencing Python in install steps, always use python3

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