Agent skill

Orchestrate

by tetherto in tetherto/qvac

Run the full implement → test → CI → review → PR pipeline for a task.

Apache-2.0Auto-check passedDevelopment

Install Orchestrate

skills CLI
$ npx skills add tetherto/qvac --skill orchestrate -a claude-code

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

GitHub CLI
$ gh skill install tetherto/qvac orchestrate --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/tetherto/qvac.git skills-src && mkdir -p .claude/skills && cp -r skills-src/packages/ocr-ggml/.agent/skills/orchestrate .claude/skills/orchestrate && 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
orchestrate
GitHub stars
681
Token cost
~2k tokens
SKILL.md length
950 words
Files
1
Skills in repo
50
Repo updated
First seen
Licence
Apache-2.0

At a glance

Run the full implement → test → CI → review → PR pipeline for a task.

  • Works in 10 steps: Setup → 5: Plan → Implement → …
  • Tasks that involve Pull requests
  • SKILL.md covers Usage, Pipeline, Error handling and Important notes
  • Calls git and gh

What it does

Orchestrate is an agent skill from tetherto/qvac. Run the full implement → test → CI → review → PR pipeline for a task. Coordinates implementer, test-writer, ci-validator, and code-reviewer agents.

Its SKILL.md is about 2k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Development, covering Pull requests and Code review. It works with Asana and Git. The repository describes itself as: Open-source local AI SDK - run AI on-device with no cloud, no API keys. Supports GGUF, RAG, image, music, and video generation, speech-to-text, P2P inference, and more… The licence is Apache-2.0.

When your agent uses it

  • Tasks that involve Pull requests
  • Tasks that involve Code review

Example prompts

  • “/orchestrate”

Workflow steps

10 steps, taken from the step headings in SKILL.md.

  1. Setup
  2. 5: Plan
  3. Implement
  4. 5: Determine test and CI requirements
  5. 75: Write Tests
  6. CI Validation
  7. Review
  8. Re-validate (if reviewer made fixes)
  9. Push and Create PR
  10. Report

What it can do on your machine

Read from SKILL.md and the folder at commit c3a6030. 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

    Shell commands in SKILL.md call:

    • git
    • gh

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

  • Network

    No URLs in SKILL.md. Its commands use git and gh, 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

Orchestrate loads about 2k tokens when it runs. Until then it costs about 40 tokens; SKILL.md has 950 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~40
When it runs · the whole SKILL.md, loaded when a task matches
~2k

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); files beside SKILL.md are not scanned.

SKILL.md

The full file from tetherto/qvac at commit c3a6030, republished under its Apache-2.0 licence (© tetherto). 950 words, ~1,998 tokens.

Download SKILL.mdSave it as .claude/skills/orchestrate/SKILL.md (or your agent's skills folder).
name
orchestrate
description
Run the full implement → test → CI → review → PR pipeline for a task. Coordinates implementer, test-writer, ci-validator, and code-reviewer agents.
argument-hint
<asana-task-id-or-url>
disable-model-invocation
true

Orchestrate: Full Task Pipeline

Run the complete agent pipeline for a task: branch setup, implement, test, CI validate, review, push, and create PR.

Usage

/orchestrate <asana-task-id-or-url>

Accepts either:

  • Asana task ID: 1213560067347874
  • Asana URL: https://app.asana.com/0/1234567890/1213560067347874

Pipeline

Phase 0: Setup
  1. Parse the Asana task ID from the argument:

    • If it's a URL, extract the last numeric segment as the task ID
    • If it's already a numeric ID, use it directly
  2. Read the Asana task to get:

    • Task title
    • Task description and acceptance criteria
    • Any tags or custom fields (e.g., package name, ticket number)
  3. Create a feature branch from main:

    • Pull latest main: git checkout main && git pull origin main
    • Generate branch name from task: feat/<ticket>-<slug>
      • Extract ticket number from task name or custom field (e.g., QVAC-123)
      • Slugify the task title (lowercase, hyphens, max 50 chars)
      • Example: feat/QVAC-123-add-rag-support-for-lancedb
    • If no ticket number found: feat/<task-title-slug>
    • Create and switch to branch: git checkout -b <branch-name>
  4. Inform the user of the setup:

    Task: <task-title>
    Branch: <branch-name>
Phase 0.5: Plan

Before any implementation, create a plan for the user to approve:

  1. Read the relevant source files mentioned in the task description or likely affected by the changes
  2. Draft an implementation plan that includes:
    • Summary of what will be changed and why
    • Files to create or modify (with brief description of changes per file)
    • Approach and key design decisions
    • Dependencies or packages to add (if any)
    • How it will be verified (build commands, test commands)
  3. Present the plan to the user and wait for approval
    • If the user requests changes, update the plan and ask again
    • Do NOT proceed until the user explicitly approves
  4. Comment on the Asana task with the approved plan
Phase 1: Implement

Launch the implementer agent with the Asana task ID and the approved plan.

Implement Asana task <task-id>. Follow this approved plan:

<paste the approved plan here>

Write code within scope of the plan, verify build/tests pass, and commit working changes.

Wait for completion. If the implementer reports failure (e.g., ambiguous requirements, build failures after 3 retries), stop the pipeline and report to the user.

Phase 1.5: Determine test and CI requirements

After implementation, analyze the changed files and the Asana task to decide what's needed next.

Run git diff --name-only main...HEAD and apply these rules:

  1. Native addon packages — if changed files match a package in the CI Package Mapping table in .agent/knowledge/ci-validation.md, CI is needed. Use the short name from that table. If multiple addon packages changed, run CI for each.

  2. SDK / TS packages (packages/sdk/**, packages/rag/**, packages/cli/**) — SDK CI runs automatically via pr-checks-sdk-pod on PR creation. No manual trigger needed.

  3. Everything else (simple libraries, docs, workflows, config, markdown) — no CI needed.

If CI is needed, inform the user which packages will be validated and why.

Determine if new tests are needed by checking:

SignalTests needed?
New public API / exported functions addedYes
New feature with user-facing behaviorYes
Bug fix (regression test)Yes
Asana task acceptance criteria mention testable behaviorYes
Refactoring with no behavior changeNo
Documentation / config / CI workflow onlyNo
Changes already have corresponding test updates from implementerNo — skip

Read the Asana task acceptance criteria. If they describe specific behaviors or scenarios, those should become tests.

Phase 1.75: Write Tests

If Phase 1.5 determined tests are needed, launch the test-writer agent:

Write automated tests for the changes on the current branch. Task ID: <task-id>. Focus on new public APIs, new behavior, and edge cases. Match existing test patterns.

Wait for completion. If the test-writer discovers code bugs, launch the implementer again with the bug details before proceeding.

If tests are not needed, skip to Phase 2.

Show full SKILL.md (403 more words)Show less
Phase 2: CI Validation

If Phase 1.5 determined CI is needed:

  1. Push the current branch: git push -u origin HEAD
  2. Launch the ci-validator agent for each affected package
  3. If CI fails with code errors: go back to Phase 1 — launch implementer again with the error details
  4. If CI fails with infra errors: let ci-validator handle retries
  5. Maximum 2 implement→CI loops before stopping

If CI is not needed, skip to Phase 3.

Phase 3: Review

Launch the code-reviewer agent:

Review all changes on the current branch against main. Task ID: <task-id>. Check requirements match, bugs, conventions, security, scope, and test coverage. Fix issues directly and commit fixes.

Wait for completion. Collect the review summary.

Phase 4: Re-validate (if reviewer made fixes)

If the reviewer committed any fixes AND Phase 1.5 determined CI was needed:

  1. Re-run CI validation for the affected packages
  2. If CI passes, proceed to reporting

If no CI needed or no reviewer fixes, proceed to reporting.

Phase 5: Push and Create PR
  1. Push the branch to origin:

    bash
    git push -u origin HEAD
  2. Determine PR type from the changed files:

    • If changes are in packages/sdk/ or other TS packages → SDK PR
    • If changes are in native addon packages → Addon PR
    • If mixed → use addon format (more detailed)
  3. Create the PR using gh pr create:

    • Title: <ticket> <prefix>[tags]: <task-title-summary> (following commit format from CLAUDE.md)
    • Body: Generate based on PR type:
      • For addon packages: follow the format from /qv-addon-pr-create
      • For SDK packages: follow the format from /qv-sdk-pr-create
      • Include: what changed, why, test plan, link to Asana task
    • Base: main
    • Example:
      bash
      gh pr create --base main --title "QVAC-123 feat: add RAG support" --body "..."
  4. Link the PR to the Asana task: comment on the task with the PR URL.

Phase 6: Report

Produce a final summary:

Pipeline complete for task <task-id>:

Branch: <branch-name>
PR: <pr-url>

Implementation:
  - [summary from implementer]
  - Files changed: [list]

Tests:
  - [added/skipped, with reason]
  - Tests added: [count and brief descriptions]
  - Code bugs found by tests: [count or none]

CI Validation:
  - [pass/fail/skipped]
  - Packages tested: [list or "n/a — no native addon changes"]
  - Platforms: [list]

Review:
  - Issues found and fixed: [count]
  - Issues flagged but not fixed: [count, with details]

Status: [ready for human review / needs attention]

Update the Asana task:

  • Add the final summary as a comment
  • If all phases passed, mark the task as complete

Error handling

  • If implementer fails: report what went wrong and stop
  • If CI fails after 2 implement→CI loops: report the persistent failure and stop
  • If reviewer finds architectural concerns: report them and stop
  • If PR creation fails: report the error, the branch is still pushed
  • At any stop point, comment on the Asana task with current status

Important notes

  • Phase 0 creates the branch automatically — user does not need to set up anything
  • The skill asks for confirmation before marking the Asana task complete
  • Each agent runs in isolation with fresh context
  • The pipeline can be resumed manually if interrupted — just re-run from the failed phase

© tetherto, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in packages/ocr-ggml/.agent/skills/orchestrate of tetherto/qvac.

Open the folder on GitHubat commit c3a6030

Compare with similar skills

Orchestrate 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.

Orchestrate compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Orchestrate this skilltetherto/qvac681—~2kAutomated safety check: PassApache-2.0
Understand Diff AnalysisEgonex-AI/Understand-Anything86k1 repos~1.4kAutomated safety check: PassMIT
Open Code Review CLIalibaba/open-code-review44k—~3.1kAutomated safety check: PassApache-2.0
Code Reviewflutter/flutter179k—~1.4kAutomated safety check: PassBSD-3-Clause
PR Review State Fetchprisma/orm48k—~767Automated safety check: PassApache-2.0
Knowledge Graph PR Reviewtirth8205/code-review-graph32k—~452Automated safety check: PassMIT

Similar skills

  • Understand Diff Analysis

    Egonex-AI/Understand-Anything

    Reads your git changes or a pull request against a prebuilt knowledge graph of the project to explain what changed, which components are affected and what is risky.

    86k GitHub starsUsed in 1 repo~1.4k tokens
    DevelopmentAuto-check passed
  • Open Code Review CLI

    alibaba/open-code-review

    Runs the ocr command-line tool to review Git changes, a commit or a branch comparison with an AI model, returning line-level comments and optionally applying fixes.

    44k GitHub stars~3.1k tokensUpdated 3 days ago
    DevelopmentAuto-check passed
  • Code Review

    flutter/flutter

    Performs a comprehensive, multi-step code review of pull requests or local code changes, using iterative refinement (generation, critique, synthesis) to ensure high-quality, actionable feedback.

    179k GitHub stars~1.4k tokensUpdated today
    DevelopmentAuto-check passed
  • Official

    Fetches a pull request's canonical review state as JSON, validates it, and renders markdown, a text summary and triage target files from it using bundled scripts.

    48k GitHub stars~767 tokensUpdated today
    DevelopmentAuto-check passed
  • Knowledge Graph PR Review

    tirth8205/code-review-graph

    Reviews a pull request or branch diff with a code knowledge graph and produces a structured review that includes blast-radius analysis.

    32k GitHub stars~452 tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Official

    Runs the triage step of the review-framework loop: reads fetched PR review state, builds `review-actions.json`, validates it and renders `review-actions.md`.

    48k GitHub stars~995 tokensUpdated today
    DevelopmentAuto-check passed

More from tetherto/qvac

All 50 skills in this repo
  • Creates a Solutions page in the QVAC documentation website from a real use case, generalizing the case into reusable guidance and registering the page in the site navigation.

    681 GitHub stars~2.8k tokensUpdated today
    Auto-check passed
  • Qv Docs Update

    tetherto/qvac

    Updates the docs website after a change to the SDK or CLI. An agent skill from tetherto/qvac.

    681 GitHub stars~11k tokensUpdated today
    Auto-check passed
  • Qv Agent Stack Sync

    tetherto/qvac

    Plan and prepare the QVAC agent-stack release cascade across @qvac/inference, @qvac/sdk, @qvac/cli, @qvac/ai-sdk-provider, @qvac/opencode-plugin, and @qvac/openclaw-plugin.

    681 GitHub stars~2k tokensUpdated today
    Auto-check passed
  • Run the deterministic code-quality audit, turn related findings into contextual remediation groups, prepare approval-gated Asana proposals, reconcile recurring runs, or configure twice-monthly…

    681 GitHub stars~1.6k tokensUpdated today
    Auto-check passed
  • Review C++ changes for string parameter and call-site efficiency conventions (std::stringview, std::string&&, const std::string&, const char, and TransparentStringMap lookup).

    681 GitHub stars~702 tokensUpdated today
    Auto-check passed
  • Qv Addon Changelog

    tetherto/qvac

    Generate changelog entries for a target add-on package. An agent skill from tetherto/qvac.

    681 GitHub stars~1.7k tokensUpdated today
    Auto-check passed

Works with

Questions about Orchestrate

What does Orchestrate do?

Run the full implement → test → CI → review → PR pipeline for a task. Orchestrate is an agent skill from tetherto/qvac. Run the full implement → test → CI → review → PR pipeline for a task.

When should I use Orchestrate?

Orchestrate fits situations like: tasks that involve Pull requests; tasks that involve Code review.

How do I install Orchestrate in Claude Code?

Run `npx skills add tetherto/qvac --skill orchestrate -a claude-code`. Or copy the skill folder (packages/ocr-ggml/.agent/skills/orchestrate in tetherto/qvac) into .claude/skills/orchestrate in your project. Claude Code loads it when a task matches its description.

How do I install Orchestrate in Codex?

Run `npx skills add tetherto/qvac --skill orchestrate -a codex`. Or copy the skill folder (packages/ocr-ggml/.agent/skills/orchestrate in tetherto/qvac) into .agents/skills/orchestrate in your project. Codex loads it when a task matches its description.

Can I use Orchestrate 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 tetherto/qvac --skill orchestrate -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/orchestrate, .gemini/skills/orchestrate, .github/skills/orchestrate and .opencode/skills/orchestrate in your project.

What does Orchestrate need to run?

Going by SKILL.md and its folder, Orchestrate needs the command-line tools its instructions call (git and gh).

Does Orchestrate access the network?

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

Is Orchestrate 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. Review the folder before installing.

What licence does Orchestrate use?

Orchestrate is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Orchestrate use?

About 2k tokens (SKILL.md is roughly 8k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Orchestrate?

Skills that share tags, products or a category with Orchestrate: Understand Diff Analysis (Egonex-AI/Understand-Anything, 86k stars), Open Code Review CLI (alibaba/open-code-review, 44k stars), Code Review (flutter/flutter, 179k stars) and PR Review State Fetch (prisma/orm, 48k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Orchestrate?

tetherto (a GitHub organization) maintains it in tetherto/qvac, which has 681 GitHub stars. The repository holds 50 skills in this directory. The repository was last updated on October 7, 2026.

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