feature-ship-build
Skill0 starsChanged 4 months agoMITView on GitHub
Phase 2 of 4 in the feature-ship chain (spec → build → review → ship). Implements the spec from Phase 1, preferring /ultraplan (isolated git worktree) or fairy-tales build-role subagents, with the post-edit .pi/test-command hook self-repairing test failures as it goes. Trigger when the previous phase says run feature-ship-build next, or standalone via "feature-ship-build", "phase 2 of feature-ship", "implement the spec", "build the feature".
In this repository21 other files it gives its agents.
- bughunt-diagnoseSkill
- bughunt-fixSkill
- bughunt-reproSkill
- bughunt-verifySkill
- feature-ship-reviewSkill
- feature-ship-shipSkill
- feature-ship-specSkill
- migrate-inventorySkill
- migrate-transformSkill
- migrate-verifySkill
- onboard-deepenSkill
- onboard-digestSkill
- onboard-mapSkill
- release-bumpSkill
- release-changelogSkill
- release-publishSkill
- release-tagSkill
- release-verifySkill
- research-artifactSkill
- research-gatherSkill
- research-synthesizeSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 3 are distinctive, and 6 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
6
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 29
Call the chain tool with action: "status", chain: "feature-ship" (a legacy .feature-ship-state.md from an older version is imported …
- Very specific: line 32
If there is no run → abort: "No feature-ship run found.
- Very specific: line 33
If the run is not active with currentPhase: "build" → abort: "Expected the run to be at phase build but it is at <currentPhase/status>.
- Very specific: line 36
If the spec content (data.spec / data.acceptanceCriteria or the spec phase summary) is missing → abort: "Phase pointer looks right but the …
- Very specific: line 39
If the tool reports the chain is locked by another session, abort.
- Very specific: line 48
If the spec's "Files to Touch" spans multiple modules, is a multi-step change, or the user wants an isolated, inspectable working copy, run …
- Distinctive: line 83
Call the chain tool:
- Distinctive: line 97
After completing this phase, output exactly:
- Distinctive: line 102
Do not start the next phase.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Feature Ship: Build
Chain version: 2 (durable state contract — see docs/STATE-CONTRACT.md)
Phase 2 of 4: feature-ship-spec → feature-ship-build → feature-ship-review → feature-ship-ship
Overview
Implements the spec produced in Phase 1. Prefers /ultraplan when the change is large
enough to warrant an isolated git worktree, or fairy-tales build-role work
otherwise, so implementation happens against a dedicated execution context rather than
inline improvisation. The post-edit .pi/test-command hook fires after every edit and
self-repairs test failures as they happen, so this phase doesn't need a separate manual
test-fix loop.
Load State
Call the chain tool with action: "status", chain: "feature-ship" (a legacy
.feature-ship-state.md from an older version is imported automatically):
- If there is no run → abort: "No feature-ship run found. Run
feature-ship-specfirst." - If the run is not
activewithcurrentPhase: "build"→ abort: "Expected the run to be at phasebuildbut it is at<currentPhase/status>. Run that phase's skill instead." - If the spec content (
data.spec/data.acceptanceCriteriaor the spec phase summary) is missing → abort: "Phase pointer looks right but the Phase 1 output is missing. Treat as corrupted — restart the chain or repair via chain action 'update'." - If the tool reports the chain is locked by another session, abort; a dead session's
lock can be cleared with
action: "unlock". Never create lock files yourself.
Note the run's runId — it keys this phase's quest dedupe keys.
Workflow
Step 1 — Choose the execution path
- If the spec's "Files to Touch" spans multiple modules, is a multi-step change, or the
user wants an isolated, inspectable working copy, run
/ultraplanagainst the spec — it opens an isolated git worktree so the implementation can be built, tested, and discarded without touching the working tree the user is looking at. - Otherwise, dispatch the implementation as fairy-tales build-role work — one unit
per independent piece of the spec, each briefed with the relevant slice of the spec
(goal, approach, files to touch, acceptance criteria) rather than the whole spec dump.
For each unit, use the quest tool so the work survives a session restart:
enqueuewithrole: "build",dedupeKey: "feature-ship/<runId>/build/<unit>",chain: { chain: "feature-ship", runId: "<runId>", phase: "build" }, andretainUntilConsumed: true; if the returned quest is alreadydone(crash-resume), reuse its result, elserunit byid;consumeeach quest after recording its output. Idempotent dedupe keys are what prevent a re-run of this phase from duplicating implementation work.
Step 2 — Implement against the acceptance criteria
Implement every file change called out in the spec. After each edit, the project's
post-edit .pi/test-command hook (if configured) runs automatically and attempts to
self-repair any test failure it introduces — do not manually re-run the same fix twice
if the hook already retried; read its output and address the root cause if it still
fails after the hook's attempt.
Step 3 — Record what was built
Capture, precisely:
- Branch name and/or worktree path (if
/ultraplanwas used). - The actual files changed (from
git status/git diff --stat), not just the files the spec predicted. - Whether a patch file was produced instead of a live branch (e.g., worktree discarded
after
/ultraplanexports a diff) — if so, its path.
Step 4 — Complete the phase
Call the chain tool:
action: "complete-phase", chain: "feature-ship", phase: "build",
summary: "implemented via <ultraplan worktree | build-role quest(s)> at <branch/worktree/patch path>; deviations from the spec (if any) and why",
data: { "filesChanged": ["<path>", "..."] },
artifacts: { "branch": "<branch or worktree path>", "buildQuests": "<quest ids>" }The tool validates this is the current phase, advances the run to review, and
rewrites the JSON + markdown projection atomically.
Handoff
After completing this phase, output exactly:
Phase 2 complete. Chain state updated (
.pi/fairy-tales/chains/feature-ship/state.json). Run/feature-ship-reviewto begin Phase 3.
Do not start the next phase. Do not offer to continue. Output only the handoff line and stop.
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
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.

