Skip to content
Skills Library
Programming · Building & coding

implement-issue

/implement-issue

Take one GitHub issue of this repo from its acceptance criteria to a verified change: read the issue, mark assumptions, plan along the chain (model, process API, domain and application, external-task workers, REST and persistence, contract, Bruno), stop for the user's approval of the plan, implement test-first, verify and report each acceptance criterion with its evidence. Use when the user says implement, build or work on issue #n. Never commits, pushes, opens a PR or changes the issue unless asked.

Licence
MIT
Last repository change
07.11.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
2
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 07.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • SKILL.md zawiera komentarze HTML (2). Nie widać ich w podglądzie Markdown, a model je czyta.

What it does and when to use it

Take one GitHub issue of this repo from its acceptance criteria to a verified change: read the issue, mark assumptions, plan along the chain (model, process API, domain and application, external-task workers, REST and persistence, contract, Bruno), stop for the user's approval of the plan, implement test-first, verify and report each acceptance criterion with its evidence. Use when the user says implement, build or work on issue #n. Never commits, pushes, opens a PR or changes the issue unless asked.

Arguments

/implement-issue <issue number or URL>
Original author description

Take one GitHub issue of this repo from its acceptance criteria to a verified change: read the issue, mark assumptions, plan along the chain (model, process API, domain and application, external-task workers, REST and persistence, contract, Bruno), stop for the user's approval of the plan, implement test-first, verify and report each acceptance criterion with its evidence. Use when the user says implement, build or work on issue #n. Never commits, pushes, opens a PR or changes the issue unless asked.

How it works

  1. 1. Read the issue
  2. 2. Separate facts from assumptions
  3. 3. Plan, then stop
  4. 4. Implement test-first
  5. 5. Verify
  6. 6. Report
  7. Only when asked
Original SKILL.md headings · 80 lines, 568 words
SKILL.md preview
---
name: implement-issue
argument-hint: "<issue number or URL>"
description: "Take one GitHub issue of this repo from its acceptance criteria to a verified change: read the issue, mark assumptions, plan along the chain (model, process API, domain and application, external-task workers, REST and persistence, contract, Bruno), stop for the user's approval of the plan, implement test-first, verify and report each acceptance criterion with its evidence. Use when the user says implement, build or work on issue #n. Never commits, pushes, opens a PR or changes the issue unless asked."
allowed-tools: Read, Write, Edit, Glob, Grep, Bash, Skill, Agent
---

# Skill: implement-issue

The human owns the requirement and the plan; the agent owns the legwork. Two stops are mandatory:
the **plan approval** (step 3) and any **business decision** the issue leaves open.

## 1. Read the issue

```bash
gh issue view <n>
gh issue view <n> --json comments --jq '.comments[].body'
```

Collect the acceptance criteria, the elements of the models it touches and linked issues. An open
issue this one depends on blocks the work: say so and stop.

## 2. Separate facts from assumptions

List every detail the plan needs. Each one is **derived** (cite the issue, the model or the code) or
an **assumption**. Business assumptions (deadlines, gateway conditions, error paths, who works a user
task) become questions, not choices.

## 3. Plan, then stop

If the issue changes a model, run the agent `review-process` on the current models now and restate
the change in business terms (which element, which path, which outcome) with the business questions;
the model edit comes after the approval.

Write the plan as a checklist along the chain, skipping hops the issue does not touch:

1. Model change (skill `bpmn-cli:bpmn`, never XML by hand; one topic per new service task) and the regenerated process API
2. Domain and application: value objects, ports, services, domain rules
3. External-task workers with their unit tests (skills `bpmn-to-code:create-worker`,
   `bpmn-to-code:test-worker`), process-adapter methods for new messages or user tasks (skill
   `bpmn-to-code:create-process-adapter`) and a process test per new path (skill
   `bpmn-to-code:test-process`)
4. REST and persistence (skills `create-rest-controller`, `create-persistence-adapter`), including the
   re-exported `openapi/openapi.json` and the Flyway migration
5. A Bruno scenario for every path a client drives over REST
6. Docs: `bruno/README.md`, `docs/assets/bike-leasing.png` if the picture changed, an ADR if a decision
   is taken (skill `create-adr`)
<!-- variant:blueprint -->

Every behaviour item lands in **both** variants with equivalent tests; models, migrations and
`application.yaml` are copied byte-identically.
<!-- /variant:blueprint -->

Per item: the files, the test that proves it, the acceptance criterion it serves. Show the plan with
the open questions and **wait for approval**. Do not write code before it.

## 4. Implement test-first

One item at a time: failing test, code, green test, the targeted check from "Verify After Each Task"
in `AGENTS.md`. When the plan turns out wrong, stop and say what changed instead of drifting.

## 5. Verify

Run the checks of "Before opening a PR" in `CONTRIBUTING.md` for every variant you touched, plus
`npm run lint:bpmn` if a model changed. A targeted run proves nothing about tests you did not touch.
If the stack runs, drive the new path with Bruno and diagnose a stuck case with
`npx pulse-cli inspect --business-key <applicationId>`. Then let the agent `review-vertical-slice` walk
the feature and run skill `review-change` on the working tree (it covers staged and new files); fix
what they report. An agent that is not loaded in the session is applied by reading its file in
`.claude/agents/`.

## 6. Report

| Acceptance criterion | Evidence (test, check, Bruno run) | Status |
|---|---|---|

Then: files changed, checks run with their result, assumptions made, what was not verified and why.

## Only when asked

- **Commit:** stage files by name (never `git add -A`), Conventional Commit, `Closes #<n>` in the
  body. Never stage `.context/` or local settings.
- **PR:** `gh pr create --base main`, title = the commit subject, body: what, why, how it was verified,
  `Closes #<n>`.
- **Issue:** comment or change labels only on request.

Versions

RepositoryScoreLicence
This page · Miragon/process-engine-blueprints100MIT
trinity-s-ai-dna/trinity100MIT
curie-eng/curie100Apache-2.0
julian-palomo/julianpalomo.dev100MIT
cassandragargoyle/skill-source-trust-assessment100MIT
callstackincubator/cordierite100MIT
Miragon/process-engine-blueprints100MIT
Miragon/process-engine-blueprints100MIT
Miragon/process-engine-blueprints100MIT

Database snapshot: 13.11.2026 · package version de6fc62147c5 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 84implement
    This skill should be used when the user asks to "implement an issue", "create a PR for issue", "implement issue #123", "turn this issue into code", or wants to implement a GitHub issue following an implementation plan and create a pull request.
  • 90work-on-issue
    Use when the user says "work on issue", "implement issue", "/work-on-issue", or asks to implement a GitHub issue end-to-end (not merely validate it). Takes an issue URL or number (defaults to the just-validated issue). Implements in an isolated worktree, verifies, commits,…
  • 100working-on-an-issue
    Use when asked to work on, read, or implement a GitHub issue - hands-off workflow (understand, scoped recon, verification plan, implementation plan, implement, verify, PR) with no mid-flow approval gates; assumptions are documented in the plan and PR, and developer review happens on the pull request
  • 100dx-implement-issue
    Use when someone wants an existing GitHub issue built, for example "implement issue 142", "pick up this issue", "build the story in issue 98", or when they paste an issue body and ask for the code. Reads the issue, checks it is specific enough to build, plans against its…
  • 100sl-ds-ticket-to-plan
    Turns a GitHub issue of the SL Design System (sl-design-system/components) into draft acceptance criteria for the developer to confirm, then into an implementation plan mapped to this repository's files. Use it when a design-system developer starts work on an issue, asks what an…
  • 100verify-acceptance-criteria
    Use when asked to "verify acceptance criteria", "check criteria", or "verify the implementation against the plan" during active feature development. Do NOT trigger for issue management (close issue, comment), worktree cleanup, or post-merge/post-PR operations.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.