implement-spec
SkillDev toolsimplement-spec is a skill that lets an AI agent turn a written specification and its tickets into a full pull request. The agent reads the spec and ticket graph, creates a branch and draft PR, then coordinates parallel coding subagents in separate worktrees to implement the work. It merges their output, runs code review, fixes issues, and cleans up the worktrees.
Use implement-spec in Claude, ChatGPT or Ahel Desktop
Free. Sign in, add implement-spec and connect your AI. About a minute.
Also: Claude Code · Cursor · Codex
Then ask your AI: use the implement-spec skill
Details
Instructions available. Your AI can read the instructions. Execution depends on the setup they require.
Account requirements not reviewed. Check the skill instructions before use; ahel provides instructions and does not run this skill.
No other account needed.
Have a written specification and its ticket graph available to the agent.
What your AI can do with it
- Read a specification and its ticket graph
- Create a branch and draft pull request for the whole spec
- Spawn implementer subagents in separate worktrees
- Merge subagent output into the PR branch
- Run code review and fix issues
- Clean up worktrees after the work is done
Getting started
- Have a written specification and its ticket graph available to the agent.
- Make sure the agent can create branches, open pull requests, and run subagents in separate worktrees.
- Ask the agent to use the implement-spec skill on that specification.
- Review the draft pull request the agent opens and the code review results it reports.
What this skill tells your AI
The instructions your AI receives, as published by mattpocock/skills in skills/in-progress/implement-spec/SKILL.md and read by ahel’s review.
You have been provided a spec. This spec should have tickets associated with it, describing how to implement the spec.
The goal is a PR which implements the entire spec on a single branch.
The tickets are not a list of steps. They are a task graph with blocking relationships between them. This means there is always a frontier of tickets which are ready to be grabbed.
Communication to and from subagents should be sparse. Communicate primarily through context pointers: to the spec, tickets, research notes, and previous commits. Don't duplicate information already available via pointers.
Implementer subagents should be run in the background where possible for maximum concurrency.
Steps
-
Read the spec and tickets. Read enough to understand the task graph.
-
(optional) Use an exploration subagent to conduct any exploration required by the tickets - relevant codebase files or external documentation. Ensure the exploration subagent can save files - it should save its markdown notes in a directory outside the repo, accessible by all future subagents. This lets implementer subagents focus on implementation rather than exploration.
-
Create a branch, and a draft PR. The PR should be marked as 'closing' the spec issue and tickets.
-
Use implementer subagents to implement each ticket. Each implementer subagent should work in its own worktree, on its own branch.
-
Once an implementer subagent completes, merge its work to the PR branch with a merger subagent.
-
If this changes the frontier of available tickets, kick off more implementer subagents to work on the new tickets. This allows for maximum concurrency.
-
Once all tickets are complete, run /code-review on the PR branch. Fix all issues raised by the code review in a single implementer subagent.
-
Mark the PR as ready for review.
-
Clean up all implementer subagent worktrees.
Signals
- GitHub stars
- 273k
- Forks
- 23k
- Last commit
- Sep 2026
- Hacker News mentions
- 1
Others that do the same job
Questions
- What does implement-spec do?
- It guides the agent to read a spec and its ticket graph, create a branch and draft PR implementing the whole spec, spawn implementer subagents in separate worktrees, merge their output, run code review, fix issues, and clean up worktrees.
- How does it handle parallel work?
- It spawns implementer subagents in separate worktrees so they can work concurrently, then merges their output into the PR branch.
- Does it run code review?
- Yes. After merging subagent output into the PR branch, it runs code review and fixes issues.
- What do I need before using it?
- A written specification and its ticket graph, plus an agent that can create branches, open pull requests, and run subagents in separate worktrees.
Advanced
- Item type
- skill
- Key
implement-spec-mattpocock- Source
- github.com/mattpocock/skills