Eys-55/ace-workflows / .agents/skills/setup-workflow-project
setup-workflow-project
VerifiedSet up a new workflow project in this repo by creating the JSON project state, task index, artifact folders, and project-level ECC concept notes under projects/<project-slug>/. Use when the user asks to create, initialize, or set up a workflow project.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.5 KB |
| agents/openai.yaml | config | 209 B |
SKILL.md
View on GitHub---
name: setup-workflow-project
description: Set up a new workflow project in this repo by creating the JSON project state, task index, artifact folders, and project-level ECC concept notes under projects/<project-slug>/. Use when the user asks to create, initialize, or set up a workflow project.
---
# Setup Workflow Project
Use this skill to initialize a project workspace. It creates project-level JSON
state; it does not create a task unless the user also invokes task initiation.
## Skill-First Runtime Rule
Follow root `AGENTS.md`: workflow projects are built as skill-first packages for
agent runtimes. Use `$skill-name` workflow surfaces for operator actions.
Command helpers may support validation or setup internals, but they are not the
primary skill-calling surface.
## Required Reading
Before acting:
1. Read root `AGENTS.md`.
2. Read `references/matt-pocock-skills.md`.
3. Read `.agents/skills/initiate-task/SKILL.md` for the task state contract.
4. Read `registry/agents-md.json`.
## Inputs
Require:
- `project:<slug>`
Accept:
- `name:"..."`
- `goal:"..."`
- `domain:"..."`
If `project:<slug>` is missing, ask for it and stop.
## Project Scaffold
Create:
```text
projects/<project-slug>/
AGENTS.md
project.json
tasks/
index.json
artifacts/
prds/
issues/
reviews/
handoffs/
```
Do not overwrite existing project files. If the project already exists, load it,
report its state, and hand off to `$review-project-state` before changing it.
Do not modify an existing project from this skill unless the user explicitly
asks after reviewing the project state.
Register `projects/<project-slug>/AGENTS.md` in `registry/agents-md.json` with:
```json
{
"path": "projects/health/AGENTS.md",
"role": "project-domain",
"scope": "projects/health",
"owning_project": "health",
"live": true,
"allowed_content": [
"project domain behavior",
"project vocabulary",
"project source rules",
"project agent workflow guidance"
]
}
```
Project `AGENTS.md` is domain-only. Root `AGENTS.md` owns workflow/task
mechanics.
## JSON Contract
Create `project.json`:
```json
{
"project_slug": "health",
"name": "Health",
"project_state": "active",
"goal": "",
"domain": "",
"agents_md": "projects/health/AGENTS.md",
"active_conventions": [],
"ecc_concepts_applied": [
"workflow contract",
"human boundary",
"project state preload"
],
"created_at": "YYYY-MM-DD",
"updated_at": "YYYY-MM-DD"
}
```
Create `AGENTS.md`:
```markdown
# Health Project AGENTS.md
This file contains live project/domain instructions for the `health` workflow
project only.
Root `AGENTS.md` owns task tracking, JSON state, Matt Pocock flow, ECC workflow
rules, validation, and repo-wide safety. Do not redefine those mechanics here.
Use this file for:
- domain vocabulary
- project-specific source rules
- project-specific output expectations
- agent behavior for this project domain
```
Create `tasks/index.json`:
```json
{
"project_slug": "health",
"tasks": []
}
```
## Output Contract
Report:
```text
PROJECT
- slug
- project_state
- files created or already present
- registered AGENTS.md path
ECC CONCEPTS
- concepts recorded
NEXT ACTION
- initiate a task with $initiate-task
```
For a newly created project, the next surface is:
```text
$initiate-task project:<project-slug> title:"..."
```
For an existing project, the next surface is:
```text
$review-project-state project:<project-slug>
```
## Developer Verification
Run:
```bash
node scripts/validate-workflow-state.mjs
```
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies. | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |
| meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要". | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |