Agent skill

Cy Loop Tasks

by compozy in compozy/compozy

Run a requested Compozy spec delivery loop through task checkpoints, scoped QA, review, and PR CI.

MITAuto-check passed

Install Cy Loop Tasks

skills CLI
$ npx skills add compozy/compozy --skill cy-loop-tasks -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install compozy/compozy cy-loop-tasks --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/compozy/compozy.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/cy-loop-tasks .claude/skills/cy-loop-tasks && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
cy-loop-tasks
GitHub stars
2.8k
Token cost
~1.4k tokens
SKILL.md length
674 words
Files
25 (incl. scripts, references, assets)
Skills in repo
47
Repo updated
First seen
Licence
MIT

At a glance

Run a requested Compozy spec delivery loop through task checkpoints, scoped QA, review, and PR CI.

  • SKILL.md covers Entry and routing, Evidence ownership, State, memory, and delegation and Failures and completion
  • Calls python3, make and git

What it does

Cy Loop Tasks is an agent skill from compozy/compozy. Run a requested Compozy spec delivery loop through task checkpoints, scoped QA, review, and PR CI.

Its SKILL.md is about 1.4k tokens, which your agent loads only when the skill is triggered. The skill folder holds 26 other files, including scripts, reference files and assets (for example `assets/iteration-summary.template.md`, `assets/state.template.yaml` and `references/checklist.md`).

The repository describes itself as: An operating system for AI agents. Plug in the agent CLIs you already use (Claude Code, Codex, Gemini CLI, Cursor) and they become a team: they split the work, hand tasks to each… The licence is MIT.

Example prompts

  • “/cy-loop-tasks”

Requirements

  • Python 3

What it can do on your machine

Read from SKILL.md and the folder at commit c46be43. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships 1 file in scripts/, which the agent can run.

    Shell commands in SKILL.md call:

    • python3
    • make
    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Cy Loop Tasks loads about 1.4k tokens when it runs, and up to ~14k if it reads all its reference files. Until then it costs about 28 tokens; SKILL.md has 674 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~28
When it runs · the whole SKILL.md, loaded when a task matches
~1.4k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~14k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from compozy/compozy at commit c46be43, republished under its MIT licence (© compozy). 674 words, ~1,402 tokens.

Download SKILL.mdSave it as .claude/skills/cy-loop-tasks/SKILL.md (or your agent's skills folder). This skill also uses 24 other files; get the full folder from GitHub.
name
cy-loop-tasks
description
Run a requested Compozy spec delivery loop through task checkpoints, scoped QA, review, and PR CI.

Loop Tasks Driver

Run a requested full spec delivery workflow as a self-healing continue loop. Keep working through implementation, applicable QA, review, and required CI until the delivery contract is met or a concrete external dependency blocks progress. Ordinary edits and single tasks do not enter this workflow.

Entry and routing

Resolve the repo root, .compozy/tasks/<slug>/_spec.md, and the user's goal once. The initializer detects tasks mode from _tasks.md plus task_*.md; otherwise it uses free mode. Optional --frontend <claude|cursor> delegates frontend implementation; --stacked publishes task checkpoints as stacked PRs. Read references/stacked-prs.md before using that option. Invocation examples are in references/goal-header-template.md when needed.

Run python3 .agents/skills/cy-loop-tasks/scripts/detect-phase.py <slug> and load only the matching procedure below. Reuse a loaded procedure while current; entering another task does not require reloading every skill or shared contract.

Detector outputProcedure to read
phase=0 action=bootstrapBootstrap
phase=B action=execute_taskTask checkpoint
phase=B action=execute_free_sliceFree-mode slice
phase=C action=qa_report or qa_executionQA, only the printed action
phase=D action=peer_reviewReview round
phase=E action=await_ci or doneDelivery

Each action completes its owned work, records meaningful evidence in memory, advances writer-owned state, and emits assets/iteration-summary.template.md. Then run detect again and continue in the same turn. The phase boundary is a resume checkpoint, not a request for user review or a reason to end the session.

Evidence ownership

Apply cy-final-verify within the existing validation step. Reuse valid evidence for unchanged source, dependencies, build/config, and environment; a handoff, phase transition, new message, or commit alone does not invalidate it.

BoundaryEvidence it owns
Task/sliceChecks proving its outcome; explicit task-owned live/visual acceptance. Record remaining integration journeys/visual rows and their owner.
Integration QARemaining changed/cross-task journeys and required visual rows on the integrated build, reusing current task evidence.
RepairReproduction and checks/journeys/visual rows invalidated by the fix.
Commit/pushRepository make gate policy with its fingerprint cache and applicable docs-only exemption.
PR deliveryRequested outcomes, resolved QA/visual obligations, review SHIP, and required CI green at the current head.

Changed user journeys need real-run evidence. Unit results alone do not close integration QA. A UI path does not require both full E2E suites, a new lab, or a complete screenshot bundle per task. With named visual references, inspect a representative state early and complete required rows at their assigned boundary. Preserve explicit acceptance requirements; route older generic per-task QA boilerplate through the ownership map without narrowing the accepted outcome.

Show full SKILL.md (284 more words)Show less

State, memory, and delegation

  • Use init-state.py and update-state.py under this skill's scripts/ as the only state writers. Task frontmatter owns task status; _tasks.md owns topology. Keep the bootstrap goal_signature unchanged; later scope decisions go in memory.
  • Load references/memory-protocol.md on entry or recovery for memory paths and format. Record meaningful decisions/evidence before completion status; preserve valid grounding and update only the current task and shared facts it changes.
  • When the detector selects a frontend lane, or QA planning is explicitly delegated/usefully independent, read the applicable lane in references/herdr-delegation.md. The controller owns state and commits; workers own their assigned diff/evidence. Reuse workers for related follow-ups.
  • Before commit-checkpoint.py, inspect staged/unstaged paths. Its git add -A contract is usable only when every dirty path belongs to this checkpoint. Otherwise stage and commit only owned paths with the checkpoint message.

Failures and completion

Repair failed required checks within the current action. Do not write final iteration state or print a completion summary for an intermediate failure. Expected search misses and normal live-worker wait expiry are not failed checks. Use references/recovery-loop.md for diagnosis, helper failures, and the external blocker test; reuse the procedure once loaded. Routine safe fixes follow existing user authorization. Missing permission or a product decision without a safe default remains a real blocker; continue independent work before asking.

For unexpected detector output/state drift, consult references/phase-transitions.md and references/state-schema.md; use references/checklist.md as a targeted resume audit, not another execution gate.

Only Phase E is successful completion: QA obligations are resolved, review is SHIP, current local evidence is PASS, and every required PR check is green at the current head. Emit assets/done-signature.txt only then. Pending/red CI remains in progress. A proven external blocker is recorded with evidence and no signature.

© compozy, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 24 other files (scripts, references, assets) in .agents/skills/cy-loop-tasks of compozy/compozy.

  • SKILL.md
  • assets/done-signature.txt
  • assets/iteration-summary.template.md
  • assets/state.template.yaml
  • references/checklist.md
  • references/goal-header-template.md
  • references/herdr-delegation.md
  • references/memory-protocol.md
  • references/phase-bootstrap.md
  • references/phase-delivery.md
  • references/phase-qa.md
  • references/phase-review.md
  • references/phase-slice.md
  • references/phase-task.md
  • references/phase-transitions.md
  • references/recovery-loop.md
  • references/stacked-prs.md
  • references/state-schema.md
  • scripts
  • … and 6 more

Open the folder on GitHubat commit c46be43

Compare with similar skills

Cy Loop Tasks next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Cy Loop Tasks compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Cy Loop Tasks this skillcompozy/compozy2.8k—~1.4kAutomated safety check: PassMIT
Pull Requestspnpm/pnpm37k—~2.2kAutomated safety check: PassMIT
Create Pull Requestcline/cline70k1 repos~1.6kAutomated safety check: PassApache-2.0
Reviewing Pull RequestsTriliumNext/Trilium38k—~4.9kAutomated safety check: PassAGPL-3.0
Pull Requestsquarkusio/quarkus16k—~840Automated safety check: PassApache-2.0
Create Pull Request with Work Item IDmakeplane/plane60k—~824Automated safety check: PassAGPL-3.0

Similar skills

  • Pull Requests

    pnpm/pnpm

    Take a change through a pull request in the pnpm repository — opening it, then staying with it after every push until CI is green and the review round is quiet.

    37k GitHub stars~2.2k tokensUpdated today
    DevelopmentAuto-check passed
  • Opens a GitHub pull request from your current branch with the gh CLI, after reviewing the commits and diff and gathering the details the PR needs.

    70k GitHub starsUsed in 1 repo~1.6k tokens
    DevelopmentAuto-check passed
  • Reviewing Pull Requests

    TriliumNext/Trilium

    A skill your agent uses whenever a Trilium pull request is to be judged — "review PR N", "is

    38k GitHub stars~4.9k tokensUpdated today
    DevelopmentAuto-check passed
  • Pull Requests

    quarkusio/quarkus

    Rules for preparing pull requests and commits in the Quarkus project: title conventions, description format, commit hygiene, labels, and contribution policies.

    16k GitHub stars~840 tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Opens a pull request for the current branch using the repo's template, a work item ID in the title and a description filled in from the actual diff.

    60k GitHub stars~824 tokensUpdated yesterday
    DevelopmentAuto-check passed
  • HTTP Requests

    thedaviddias/Front-End-Checklist

    A skill your agent uses when auditing slow page loads, heavy assets, or rendering delays related to Minimize HTTP requests.

    74k GitHub stars~415 tokensUpdated yesterday
    Frontend & DesignAuto-check passed

More from compozy/compozy

All 47 skills in this repo
  • Eng Real Scenario QA

    compozy/compozy

    Dogfoods Compozy through an autonomous startup scenario with live providers, cross-surface observation, and strict evidence audit.

    2.8k GitHub stars~3.4k tokensUpdated today
    Auto-check passed
  • Eng Test Conventions

    compozy/compozy

    Go test-shape discipline for Compozy. An agent skill from compozy/compozy.

    2.8k GitHub stars~467 tokensUpdated today
    Auto-check passed
  • Assistant UI

    compozy/compozy

    Guide for assistant-ui library - AI chat UI components. An agent skill from compozy/compozy.

    2.8k GitHub stars~958 tokensUpdated today
    Auto-check passed
  • Guide for assistant-ui UI primitives - ThreadPrimitive, ComposerPrimitive, MessagePrimitive.

    2.8k GitHub stars~999 tokensUpdated today
    Auto-check passed
  • Assistant UI Runtime

    compozy/compozy

    Guide for assistant-ui runtime system and state management. An agent skill from compozy/compozy.

    2.8k GitHub stars~856 tokensUpdated today
    Auto-check passed
  • Assistant UI Streaming

    compozy/compozy

    Guide for assistant-stream package and streaming protocols. An agent skill from compozy/compozy.

    2.8k GitHub stars~813 tokensUpdated today
    Auto-check passed

Questions about Cy Loop Tasks

What does Cy Loop Tasks do?

Run a requested Compozy spec delivery loop through task checkpoints, scoped QA, review, and PR CI. Cy Loop Tasks is an agent skill from compozy/compozy. Run a requested Compozy spec delivery loop through task checkpoints, scoped QA, review, and PR CI.

How do I install Cy Loop Tasks in Claude Code?

Run `npx skills add compozy/compozy --skill cy-loop-tasks -a claude-code`. Or copy the skill folder (.agents/skills/cy-loop-tasks in compozy/compozy) into .claude/skills/cy-loop-tasks in your project. Claude Code loads it when a task matches its description.

How do I install Cy Loop Tasks in Codex?

Run `npx skills add compozy/compozy --skill cy-loop-tasks -a codex`. Or copy the skill folder (.agents/skills/cy-loop-tasks in compozy/compozy) into .agents/skills/cy-loop-tasks in your project. Codex loads it when a task matches its description.

Can I use Cy Loop Tasks in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add compozy/compozy --skill cy-loop-tasks -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cy-loop-tasks, .gemini/skills/cy-loop-tasks, .github/skills/cy-loop-tasks and .opencode/skills/cy-loop-tasks in your project.

What does Cy Loop Tasks need to run?

Going by SKILL.md and its folder, Cy Loop Tasks needs the command-line tools its instructions call (python3, make and git). Our summary lists: Python 3.

Does Cy Loop Tasks access the network?

SKILL.md contains no URLs. Its commands use git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Cy Loop Tasks safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.

What licence does Cy Loop Tasks use?

Cy Loop Tasks is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Cy Loop Tasks use?

About 1.4k tokens (SKILL.md is roughly 5.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 13k tokens, read only when the agent opens those files.

What are the alternatives to Cy Loop Tasks?

Skills that share tags, products or a category with Cy Loop Tasks: Pull Requests (pnpm/pnpm, 37k stars), Create Pull Request (cline/cline, 70k stars), Reviewing Pull Requests (TriliumNext/Trilium, 38k stars) and Pull Requests (quarkusio/quarkus, 16k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Cy Loop Tasks?

compozy (a GitHub organization) maintains it in compozy/compozy, which has 2,791 GitHub stars. The repository holds 47 skills in this directory. The repository was last updated on October 7, 2026.

Source: compozy/compozy on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.