agentleFS
agentdecksdkagentdecksdk/agentdeck/.claude/skills/ship-issue/SKILL.md

ship-issue

Skill4 starsChanged 2 months agoMITView on GitHub

Run one agentdeck issue through the full pipeline: brief the user, deck-dev implements it, deck-reviewer gates it, findings get fixed, PR merges to dev, board updated.

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 2 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 6 edited versions we read keep these.

  • 2

    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. Very specific:

    Update GitHub Project (PVT_kwHOBHijkM4BgHFZ): Set Status = In progress, record Start Date, comment with start timestamp.

    line 19
  2. Distinctive:

    If it stops at its own spec gate despite step 0, the issue needs a human decision: set the board back to Todo, surface what is missing, …

    line 26
  3. Very specific:

    Verify the draft PR targets dev, includes Closes #<n>, and its body carries the design sections (## Reuse analysis, ## Analog, ## Concept …

    line 40
  4. Very specific:

    Read those payloads whole.

    line 59
  5. Distinctive:

    Keep text between tool calls to ≤25 words.

    line 67

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

Ship an issue

Orchestrate the end-to-end delivery of an issue.

0. Brief Before Launching

Read the issue and verify its claims against the tree. Spec gate first: if the issue lacks Done when outcomes or scope bounds (what must NOT be added), fix the issue body or ask the user; do not spawn deck-dev against an unbounded spec. Then output a concise 5-point brief (1–2 lines each):

The claim: What the issue asks for. What's true now: Current behavior in the codebase (file:line). The shape: Implementation plan in 3–4 concise bullets. The hard gate: Invariants that must not break (tests/golden/, public API, import law). Skipped: What is deliberately out of scope.

1. Board Tracking

Update GitHub Project (PVT_kwHOBHijkM4BgHFZ): Set Status = In progress, record Start Date, comment with start timestamp.

2. Implement

Provision the worktree first: git -C <repo>/dev worktree add <repo>/wrts/<branch> <branch> (create the branch from its base if it does not exist; on a stack the base is the layer below, otherwise dev). Spawn deck-dev with the issue number, the brief, the absolute worktree path and the PR base branch. It runs Stage 0 (Understand) then Stage 1 (Design, posted to the draft PR body), then stops there and reports back rather than continuing into Stage 2.

If it stops at its own spec gate despite step 0, the issue needs a human decision: set the board back to Todo, surface what is missing, stop the pipeline.

Design gate: compare the reported design against this step's own brief. A design is never blocked for being larger; it is blocked only for a named problem: a duplication of something already in scripts/repomap.py's output, an abstraction with one caller, a divergence from its own declared ## Analog, configuration for a value that never changes. No such problem: send deck-dev a message to proceed to Stage 2. A problem found: send it back naming the problem, not a target size; deck-dev revises Stage 1 and reports again before Stage 2 starts.

Once past the gate, deck-dev continues through Implement, Self-review, and Gate, posting task progress; poll for stalls.

3. Attach PR

Verify the draft PR targets dev, includes Closes #<n>, and its body carries the design sections (## Reuse analysis, ## Analog, ## Concept budget, ## Expected delta). Missing sections go back to deck-dev before any review is spent. Set Status = In review when marked ready.

4. Review

Spawn deck-reviewer on the PR. The reviewer must post a COMMENTED GitHub review; a session-only verdict does not count.

5. Address Findings

The review lands on the PR itself (GitHub review + inline comments); read it there. Route by class: BLOCK goes to deck-dev to fix on the branch. DISCUSS is answered in the thread by the author; if the reviewer and author still disagree once answered, escalate to the user rather than letting either side rule. DEFER is already a finding:-titled issue the reviewer filed and linked; schedule it as its own harness PR, never implemented by the reviewer. NIT needs no action unless the author wants it.

6. Merge

Merge when every required check is green and every review thread is resolved: gh pr merge --squash --delete-branch.

Green checks are not sufficient. One open thread holds the PR at BLOCKED with nothing in gh pr checks to explain it, and --admin cannot force it: enforce_admins is on and the ruleset has no bypass actors. Two independent mechanisms enforce it under two different names, and both are live, so read both rather than assuming one mirrors the other:

gh api repos/{owner}/{repo}/branches/dev/protection    # required_conversation_resolution
gh api repos/{owner}/{repo}/rules/branches/dev         # required_review_thread_resolution
gh api graphql -f query='{repository(owner:"{owner}",name:"{repo}"){pullRequest(number:<n>){reviewThreads(first:20){nodes{id isResolved}}}}}'

Read those payloads whole. Filtering the protection endpoint down to the fields you expected is how this rule got missed in the first place: absence from a --jq selection is not absence from the data.

Step 6 detects; it does not resolve. The author resolves each thread, and ship-pr carries the mutation. When the orchestrator authored the PR itself, it is the author.

7. Complete

Set Status = Done, record Target Date, comment with completion summary and PR link.

Ground Rules

  • Keep text between tool calls to ≤25 words. Keep final responses to ≤100 words unless more detail is required.

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.