agentleFS
team-attentionteam-attention/hoyeon/skills/ultrawork/SKILL.md

ultrawork

Skill173 starsChanged 8 months agoMITView on GitHub

This skill should be used when the user says "/ultrawork", "ultrawork", or wants to run the full specify → execute pipeline automatically with a single command. Automated end-to-end workflow that chains specify and execute skills.

We split this skill into the rules it gives an agent

4 rules. 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 7 edited versions we read keep these.

  • 2

    Distinctive

    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
  1. Very specific:

    When specify finishes L4 with user approval → Hook triggers /execute {spec-path}

    line 29
  2. Distinctive:

    When all tasks complete → Pipeline ends

    line 31
  3. Very specific:

    Write approved spec.json at .hoyeon/specs/{name}/spec.json

    line 83
  4. Distinctive:

    After specify completes with an approved spec:

    line 87

“Common” is measured: kept by at least 6 in 10 of the 7 edited versions of this file we could read. “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
SKILL.md

/ultrawork Skill - Automated Development Pipeline

You are initiating an ultrawork session — a fully automated pipeline that chains:

  1. /specify — Systematic requirements derivation producing requirements.md
  2. /blueprint — Contract-first planning producing plan.json from requirements
  3. /execute — Reads plan.json, dispatches workers, and implements

How It Works

The ultrawork pipeline runs automatically through Stop hooks:

  • When specify finishes L4 with user approval → Hook triggers /execute {spec-path}
  • /execute then derives plan.json from the approved spec and runs tasks to completion
  • When all tasks complete → Pipeline ends

You don't need to manually trigger the next step — the hooks handle transitions.

v2 schema note

In spec-v2, spec.json contains only requirements + sub-requirements (with GWT) + verification journeys. It does not contain tasks[]. Task breakdown lives in a sibling plan.json derived by /execute. Ultrawork is just the glue: specify produces the spec, execute produces and runs the plan.

Your Role

  1. Extract the feature name from user's request
  2. Initialize ultrawork state (handled by UserPromptSubmit hook)
  3. Start the specify skill with the feature name
  4. Follow specify's layer flow normally (approve at L2, L3, L4)
  5. The rest happens automatically via hooks

Execution

Step 1: Parse User Request

Extract a short, kebab-case name for the feature:

  • "Add user authentication" → user-auth
  • "Implement payment processing" → payment-processing
  • "Fix login bug" → fix-login-bug

Note: State initialization is handled automatically by UserPromptSubmit hook (ultrawork-init-hook.sh).

Step 2: Announce Ultrawork Mode

Ultrawork Mode Activated

Feature: {name}
Pipeline: specify (spec.json v2) → execute (plan.json + run)

Starting specify L0...

Step 3: Invoke Specify

Skill("specify", args="{name}")

The specify skill will:

  1. L0: Goal mirror + confirmed_goal
  2. L1: Codebase research
  3. L2: Decisions + constraints (user approves)
  4. L3: Requirements + sub-requirements with GWT (user approves)
  5. L4: Verification journeys (user approves)
  6. Write approved spec.json at .hoyeon/specs/{name}/spec.json

Step 4: Let Hooks Handle the Rest

After specify completes with an approved spec:

  • ultrawork-stop-hook.sh detects an approved spec.json (meta.approved_by populated)
  • Hook automatically injects /execute .hoyeon/specs/{name}/spec.json
  • /execute prompts (via AskUserQuestion) for dispatch/work/verify mode selections, derives plan.json, and runs tasks to completion

User Interruption

User can stop the pipeline at any time by saying:

  • "stop"
  • "pause"
  • "wait"

This will halt the current phase and await further instructions.

State Tracking

The hook tracks progress in .hoyeon/state.local.json:

{
  "session-id": {
    "ultrawork": {
      "name": "feature-name",
      "phase": "specify",
      "iteration": 0
    }
  }
}

Phases: specify → executing → done

Example Flow

User: "/ultrawork add dark mode support"

[Hook auto-initializes state for "dark-mode"]

[You]
1. Parse: feature name = "dark-mode"

2. Announce:
   Ultrawork Mode Activated
   Feature: dark-mode
   Pipeline: specify → execute
   Starting specify L0...

3. Invoke: Skill("specify", args="dark-mode")

[Specify L0→L1→L2(approve)→L3(approve)→L4(approve)]
[spec.json written and approved at .hoyeon/specs/dark-mode/spec.json]
[Hook detects approved spec → triggers "/execute .hoyeon/specs/dark-mode/spec.json"]
[/execute prompts for dispatch/work/verify, derives plan.json, runs tasks]
[All tasks completed]
[Pipeline ends]

Important Notes

  • State is auto-initialized by UserPromptSubmit hook — no manual setup needed
  • Do NOT manually call /execute — hooks handle this
  • Follow specify's layer flow — approve at L2/L3/L4 gates
  • spec.json has no tasks in v2 — /execute derives plan.json as a sibling file
  • Mode selection is interactive — /execute asks for dispatch/work/verify via AskUserQuestion (no CLI flags)
  • The pipeline is autonomous — just start it and let it run
  • User can interrupt at any time for manual control

Download now (ZIP)

  • Other versions8 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 8

    1. Yeachan-Heo/oh-my-codex★ 34ksame name
    2. zereight/gitlab-mcp★ 2.1ksame name
    3. first-fluke/fullstack-starter★ 223same name
    4. E4crypt3d/oh-my-qwen★ 9same name
    5. ibahgat/oh-my-iflow★ 2same name
    6. Jkudjo/oh-my-cursor★ 1same name
    7. Jurel89/copilot-omni★ 0same name
    8. zephyrpersonal/oh-my-claude-code★ 0same name

    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.

  • Used alongsideProjects with it also use ralplan, cancel and 8 more.

    Other files people use alongside this one.

    Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
    • ibahgatSkill

      ralplan

      Create a strict, quality-first execution plan with explicit gates and rollback-ready sequencing.

      Also in 5 projects that use this

    • ibahgatSkill

      cancel

      Gracefully stop active autonomous workflow, preserve progress, and produce a resume-ready handoff.

      Also in 5 projects that use this

    • Jurel89Skill

      ultraqa

      QA cycling loop — build/lint/test, fix, retry up to 5 cycles

      Also in 3 projects that use this

    • Jurel89Skill

      ai-slop-cleaner

      Clean AI-generated code slop with a regression-safe, deletion-first workflow and optional reviewer-only mode

      Also in 3 projects that use this

    • klauswgSkill

      deep-interview

      在明确批准执行之前,通过数学化歧义门控进行苏格拉底式深度访谈

      Also in 4 projects that use this

    • joker-yeSkill

      git-master

      Git expert for atomic commits, rebasing, and history management with style detection

      Also in 3 projects that use this

    • +1 more from ibahgat
  • In this repository27 other files it gives its agents.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

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.