jrjsmrtnjrjsmrtn/project-orchestration-skills/CLAUDE.md

project-orchestration-skills

CLAUDE.md17 starsChanged 10 days agoMITView on GitHub

Claude Code plugin providing 12 skills for AI-assisted project orchestration. Each SKILL.md follows the Agent Skills specification with YAML frontmatter: Installation requires a marketplace — install via [jrjsmrtn-skills](https://github.com/jrjsmrtn/jrjsmrtn

Reads credentials

In this repository13 other files it gives its agents.
- analyze-projectSkill
- bootstrap-projectSkill
- graduate-backlogSkill
- harden-github-actionsSkill
- harden-gitlab-ciSkill
- plan-sprintSkill
- setup-adrsSkill
- setup-architecture-as-codeSkill
- setup-container-securitySkill
- setup-git-hooksSkill
- setup-pre-commitSkill
- validate-quality-configSkill
- wrapup-sprintSkill

The file

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

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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

Distinctive:
When writing a sibling, build that forge's equivalent controls.
line 62

≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.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

CLAUDE.md

AI-Assisted Project Orchestration Skills

Claude Code plugin providing 12 skills for AI-assisted project orchestration.

Project Structure

.claude-plugin/
  plugin.json          # Plugin manifest (name, version, description, keywords)
skills/
  <skill-name>/
    SKILL.md           # Skill definition (frontmatter + markdown body)

Skill File Format

Each SKILL.md follows the Agent Skills specification with YAML frontmatter:

---
name: <skill-name>
description: <one-line description>. <when-to-use sentences>.
license: MIT
metadata:
  author: "Georges Martin <jrjsmrtn@gmail.com>"
  version: "0.1.x"
---

name must match the directory name
description is the trigger: first sentence describes what it does, subsequent sentences describe when to use it
license is the spec-defined license field (replaces prior SPDX comment headers)
metadata.version tracks with the parent plugin's version in plugin.json
Body follows a consistent structure: title, When to Use, Required Inputs, Workflow (phased steps), Outputs (checklist), Validation, Related Skills

Skills Overview

Skill	Phase
analyze-project	Inception (SPARK methodology)
bootstrap-project	Inception (scaffolding + initial roadmap)
setup-adrs	Inception (architecture decisions)
setup-architecture-as-code	Architecture (C4/Structurizr)
setup-git-hooks	Quality (two-stage hooks)
validate-quality-config	Quality (config consistency)
harden-github-actions	Quality (GitHub Actions supply-chain hardening — deliberately forge-bound)
harden-gitlab-ci	Quality (GitLab CI supply-chain hardening — deliberately forge-bound; sibling of the above)
setup-container-security	Quality (container image hardening)
plan-sprint	Sprint workflow (planning, minimal or structured)
wrapup-sprint	Sprint workflow (retrospective, minimal or expanded)
graduate-backlog	Work organization (in-repo backlog → forge tracker, thin-hybrid)

Development Conventions

Version: 0.1.x during development (semantic versioning)
License: MIT, declared via license: MIT in every SKILL.md frontmatter (per Agent Skills spec)
Skill names are kebab-case
Skills reference each other via Related Skills sections
Skills are designed to run in the target project's working directory, not in this plugin repo
Forge-agnostic by default — a skill that touches a forge abstracts over gh / glab / tea, including self-hosted GitLab/Forgejo, and verifies command syntax against the installed CLI rather than asserting it (graduate-backlog is the reference implementation).
Sanctioned exception — the harden-* family (harden-github-actions, harden-gitlab-ci): CI hardening controls are properties of a specific CI execution model, not portable concepts wearing one forge's syntax, so each forge gets a sibling skill named for its coupling rather than a branch inside a shared skill. Siblings also keep each description a clean activation trigger. This is a closed family — one skill per forge, currently 2 of 3 (Forgejo Actions is roadmap H9). A non-harden-* exception should require an ADR.
- When writing a sibling, build that forge's equivalent controls; do not transliterate another forge's. The risks differ in kind (e.g. GitLab's CI_JOB_TOKEN defaults to own-project-only, so the work is keeping it scoped — the opposite posture from GitHub's repo-wide-by-default GITHUB_TOKEN).
Lesser forge coupling that is not sanctioned, and should be fixed if touched: bootstrap-project writes .github/dependabot.yml; wrapup-sprint calls gh api repos
The plugin is extracted from the AI-Assisted Project Orchestration pattern language project

Installation

Installation requires a marketplace — install via jrjsmrtn-skills.

Remotes

origin: private git server
github: https://github.com/jrjsmrtn/project-orchestration-skills

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