Setup Ralph.

PROMPT_build.md

@@ -0,0 +1,84 @@
+# PhaseFlow Build Mode
+
+Implement functionality from the plan, validate with tests, and commit.
+
+## 0. Orient
+
+0a. Study `specs/*` with parallel Sonnet subagents to learn the application specifications.
+0b. Study @IMPLEMENTATION_PLAN.md to understand current priorities.
+0c. Study @AGENTS.md to understand build/test commands and codebase patterns.
+0d. For reference, the application source code is in `src/*`.
+
+## 1. Select and Implement
+
+Your task is to implement functionality per the specifications using parallel subagents.
+
+Follow @IMPLEMENTATION_PLAN.md and choose the most important item to address.
+
+Before making changes, search the codebase (don't assume not implemented) using Sonnet subagents. You may use up to 8 parallel Sonnet subagents for searches/reads and only 1 Sonnet subagent for build/tests. Use Opus subagents when complex reasoning is needed (debugging, architectural decisions).
+
+## 2. TDD Workflow
+
+Follow TDD per CLAUDE.md:
+
+1. **Write a failing test** that defines the desired behavior
+2. Run the test to confirm it fails: `pnpm test:run`
+3. **Write minimal code** to make the test pass
+4. Run tests to confirm success
+5. **Refactor** while keeping tests green
+6. Repeat for each piece of functionality
+
+Ultrathink.
+
+## 3. Update Plan
+
+When you discover issues, immediately update @IMPLEMENTATION_PLAN.md with your findings using a subagent.
+
+When resolved, update and remove the item.
+
+## 4. Commit
+
+When the tests pass:
+1. Update @IMPLEMENTATION_PLAN.md
+2. Run `pnpm lint` and fix any issues
+3. `git add -A`
+4. `git commit` with a descriptive message (NEVER use --no-verify)
+5. `git push`
+
+## 5. Documentation
+
+When authoring documentation, capture the why — tests and implementation importance.
+
+## 6. Quality Rules
+
+- Single sources of truth, no migrations/adapters
+- If tests unrelated to your work fail, resolve them as part of the increment
+- Implement functionality completely. Placeholders and stubs waste efforts and time redoing the same work
+- All files must start with 2-line ABOUTME comments
+- No mock mode - use real data and APIs only
+
+## 7. Maintenance
+
+- Keep @IMPLEMENTATION_PLAN.md current with learnings using a subagent — future work depends on this to avoid duplicating efforts
+- Update especially after finishing your turn
+- When @IMPLEMENTATION_PLAN.md becomes large, periodically clean out completed items
+
+## 8. Operational Learning
+
+When you learn something new about how to run the application, update @AGENTS.md using a subagent but keep it brief. For example, if you run commands multiple times before learning the correct command, that file should be updated.
+
+## 9. Bug Handling
+
+For any bugs you notice, resolve them or document them in @IMPLEMENTATION_PLAN.md using a subagent, even if unrelated to the current piece of work.
+
+## 10. Spec Consistency
+
+If you find inconsistencies in the `specs/*` then use an Opus subagent with 'ultrathink' requested to update the specs.
+
+## 11. Keep AGENTS.md Lean
+
+IMPORTANT: Keep @AGENTS.md operational only — status updates and progress notes belong in @IMPLEMENTATION_PLAN.md. A bloated AGENTS.md pollutes every future loop's context.
+
+## 12. Exit
+
+After committing, exit cleanly. The loop will restart with fresh context.