agentleFS
usk6666usk6666/yorishiro-proxy/.claude/skills/implement/SKILL.md

implement

Skill19 starsChanged 22 days agoApache-2.0View on GitHub

Load a Linear Issue and execute the full workflow from branch creation to implementation, testing, commit, and PR creation

  • Commits and pushes

192 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use setup-ts-deep-modules, ask-matt and 10 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.
  • liufeng1976Skill

    setup-ts-deep-modules

    Wire dependency-cruiser into a TypeScript repo so each package is a deep module — implementation hidden in subfolders, reachable only through its entry-point files. User-invoked.

    Also in 43 projects that use this

  • GoogleSkill

    ask-matt

    Ask which skill or flow fits your situation. A router over the skills in this repo.

    Also in 91 projects that use this

  • RoentekSkill

    loop-me

    Grill me about specs for the workflows I want to build, within this workspace.

    Also in 50 projects that use this

  • AWSSkill

    to-questionnaire

    Turn a decision you can't fully answer into a questionnaire for someone else to fill in.

    Also in 88 projects that use this

  • ryotaroh180105Skill

    claude-handoff

    Hand the current conversation off to a fresh background agent that picks up the work immediately.

    Also in 42 projects that use this

  • GoogleSkill

    to-spec

    Turn the current conversation into a spec and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.

    Also in 168 projects that use this

  • +3 more from google

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 5 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 175 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Update Issue status: Update status to "In Progress"

    line 17
  2. Distinctive:

    Commit: Commit in Conventional Commits format, as a separate commit after the Step 6 red commit (do not amend or squash them together)

    line 31
  3. Very specific:

    Create PR with gh pr create

    line 38
  4. Distinctive:

    Update Issue status: Update status to "In Review"

    line 39
  5. Very specific:

    Read .claude/agents/design-reviewer.md and replace the placeholders before launching:

    line 44
  6. Distinctive:

    If build or tests fail, fix the issues before re-running

    line 94
  7. Very specific:

    If PR creation fails, guide the user to manually run /pr

    line 98
  8. Distinctive:

    Before completing implementation, verify that the feature is accessible from the config file (no missing config struct, validation, or init …

    line 99

“Common” is measured: kept by at least 6 in 10 of the 175 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

/implement

An end-to-end workflow skill for implementing a Linear Issue through to PR creation.

Arguments

  • /implement <Issue ID> — Implement the specified Issue (e.g., /implement USK-12)

Steps

  1. Load Issue: Fetch Issue details with mcp__linear-server__get_issue
  2. Update Issue status: Update status to "In Progress"
  3. Create branch: Generate an appropriate branch name from the Issue
    • feat: feat/<id>-<short-desc>
    • fix: fix/<id>-<short-desc>
    • other: chore/<id>-<short-desc>
  4. Design review (mandatory for any Issue that introduces a new public type, modifies an interface, or touches more than one package; skip for docstring-only / test-only / single-call-site fixes — note the skip reason in the user-facing plan): launch the design-reviewer agent. See "Design Review Invocation" below.
  5. Present plan to user: Issue summary, file actions (new / modified / untouched), key resolved decisions from the design review, the Test Plan (the failing tests that will be written first), deferred items, unresolved questions. Get user confirmation before implementing. For trivial Issues that skipped Step 4, present a one-paragraph plan that still names the failing test(s) to write first, or the TDD exemption that applies.
  6. Red — write failing tests first: Follow "TDD Workflow" Phase 1 in .claude/agents/implementer.md (it is the single source for the rules and exemptions). Write the tests plus compile-only scaffolding, run them with make test-pkg PKG=<pkg> RUN=<regexp>, confirm each fails on its assertion (not a compile error or no tests to run), and commit them alone as test(<scope>): add failing tests for <Issue ID>. Keep the --- FAIL: lines for the PR body.
  7. Green, then refactor: Implement until the Phase 1 tests pass, without weakening any Phase 1 assertion (implementer.md Phase 2), then refactor.
  8. Verify:
    • Auto-format with gofmt -w .
    • make lint (gofmt check + golangci-lint: govet / staticcheck / unused / ineffassign / gocyclo)
    • make build
    • make test
  9. Commit: Commit in Conventional Commits format, as a separate commit after the Step 6 red commit (do not amend or squash them together)
    • Include Refs: <Issue ID> in the commit message footer
  10. Push: Push to remote with git push -u origin <branch-name>
  11. Create PR:
    • Review changes with git diff main...HEAD
    • PR title in Conventional Commits format (e.g., feat(protocol): add HTTP handler)
    • PR body follows the template below
    • Create PR with gh pr create
  12. Update Issue status: Update status to "In Review"
  13. Report results: Display implementation summary + PR URL

Design Review Invocation

Read .claude/agents/design-reviewer.md and replace the placeholders before launching:

PlaceholderValue
{{SCOPE_DESCRIPTION}}Issue title + description + tentative file action list (new / modified) inferred from the Issue
{{SPEC_REFERENCES}}Paths to relevant spec/design docs (docs/rfc/envelope.md for data path; protocol-specific RFCs; docs/rfc/plugin-migration.md for plugin work). Cite specific sections, not whole files.
{{PACKAGES_TO_SURVEY}}Packages the new code creates, modifies, or depends on
{{COMPLETED_CONTEXT}}One-paragraph summary of relevant existing public surface this Issue will integrate with (types/functions, not the whole architecture)
{{PRODUCT_IDENTITY}}Read from .claude/skills/review-gate/SKILL.md Phase 1-4 "Product context" block — single source of truth
{{PRINCIPLES}}The MITM Implementation Principles from CLAUDE.md (the 6-item list). Quote verbatim.

Launch:

Agent(
  description="Design review: <Issue ID>",
  subagent_type="general-purpose",
  prompt=<composed prompt>
)

Process the result:

  • All resolved → fold the resolved decisions table into the Step 5 plan
  • Unresolved items exist → present only the unresolved questions to the user in Step 5 with the agent's proposed answers and trade-offs. Do not ask the user about resolved questions.
  • Fitness check FAIL → stop and report. Do not proceed to Step 5 until the scope is adjusted.

PR Body Template

## Summary
- <bulleted list of changes>

## Test plan
- [ ] Test items

## TDD evidence
- Red commit: <sha> (or: exempt — <reason>)
- Failing output at the red commit:

--- FAIL: TestX_Scenario ... <assertion message>

- Phase 1 test changes after the red commit: none (or: <commit> — <why>)

Resolves <Issue ID>
Linear: https://linear.app/usk6666/issue/<Issue ID>

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Notes

  • If build or tests fail, fix the issues before re-running
  • Determine implementation scope from Issue labels and description
  • The design review in Step 4 already serves as the implementation plan for non-trivial Issues — Step 5 just relays it to the user
  • Since build/test verification was done in Step 8, do not re-run when creating the PR
  • If PR creation fails, guide the user to manually run /pr
  • Before completing implementation, verify that the feature is accessible from the config file (no missing config struct, validation, or init function changes)

Download now (ZIP)

Does it work?

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

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

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.