Factory Work

Implement one task. Keep the change no larger than the task's observable outcome requires.

Establish the contract

Read .factory/CONTEXT.md, the ticket's plan.md, optional adr.md, the target task, its blockers, relevant code and tests, and project instructions. Confirm the ticket and task identifiers before editing. Do not begin when a dependency is incomplete or the worktree is shared with another active ticket.

If the plan conflicts with current evidence, update plan.md under Discoveries and adjust affected task scope before continuing. Escalate when the new evidence changes product behavior or invalidates multiple tasks.

Read references/playbooks.md and use only the playbook matching the task type.

Design before editing

Sketch the smallest viable shape:

• Start with the caller or user-visible behavior.
• Identify the highest existing test seam.
• Name the modules and public interfaces that must change.
• State what remains unchanged.
• Prefer one clear path over speculative extension points.

Keep this sketch in the session unless it changes the durable plan. Do not scaffold unused abstractions. If implementation repeatedly fights the sketch, stop, incorporate the evidence, and redesign instead of layering exceptions.

Implement in proving slices

1. Add the smallest failing test that demonstrates the next required behavior when a meaningful automated test is feasible.
2. Confirm it fails for the expected reason.
3. Write the least code that makes it pass.
4. Refactor only where the passing behavior exposes duplication or unclear ownership.
5. Repeat until the acceptance criteria are covered.

Tests should exercise public behavior, not mirror implementation. For a behavior that cannot reasonably begin with a failing test, state why and use the closest reliable validation. Do not add test-only production interfaces.

Add or update E2E coverage for user stories when the project has a viable harness. If automation is unavailable or disproportionately expensive, record the reason and a concrete human verification path in the task.

Run focused checks while working, then the relevant broader suite once. Do not spend budget repeatedly running an unchanged full suite.

Finish the work stage

Prove the real artifact before claiming the task is ready. Follow ../prove-it-works/SKILL.md. A compile, typecheck, file timestamp, or self-report is not proof. Write .factory/evidence/<ticket>/<task>/manifest.json in the factory root with one criterion per acceptance check: id, exact command, cwd, exit code, and artifact path. When the project has .factory/skills/verify-<repo>/, drive that skill and keep its screenshots or transcripts in the evidence directory.

Mark satisfied acceptance criteria and set the task status to ready_for_review, not done; review owns approval. Update durable context only for facts useful beyond this ticket.

In a supervised run, report transitions and blockers in the result and let the supervisor update state and usage. In a direct invocation, use node …/factory-supervise/scripts/fstate/cli.mjs. Never open .factory/db/ by hand.

Do not start the reviewer, commit, push, publish, or create a pull request. The supervisor owns the next stage.

Output

Return only this template (factory.work.v3). No JSON object, no recap after it. Do not list changed paths; the supervisor already has git.

## Status
ready_for_review | blocked

## Summary
One sentence for the observable result.

## Tests
- `npm test`: passed — 22 files / 109 tests
- `npm run lint`: not_run — unchanged config

## E2E
passed | failed | not_feasible
Note: evidence or human path.

## Review responses
none

- R1-01: fixed — tests/filter.test.ts:88

## Plan changed
false

## Blocker
none

## Evidence
.factory/evidence/PROJ-123/01/manifest.json

Status is ready_for_review or blocked. A blocked result must explain the blocker; it must not claim completed verification. Use none under Review responses when this is not a later round. Use none under Evidence only when the task is blocked. Each test line is command: passed|failed|not_run plus an optional note. Evidence is the manifest path written under the factory root, not the worktree.

After the template, call subagent_done and emit nothing else. caller_ping is the only other terminal call, and it replaces the template when you need a decision.