# Task 008: Boundary Questions Loop

Development description: Implement generation, editing, partial saving, and submission of boundary questions before plan generation.

## Implementation Details

- Backend endpoints:
  - `POST /api/articles/{article_id}/boundary-questions/generate`
  - `GET /api/articles/{article_id}/boundary-questions`
  - `PATCH /api/articles/{article_id}/boundary-questions/{question_id}`
  - `POST /api/articles/{article_id}/boundary-questions/submit`
- Generate 5-10 questions covering:
  - Audience.
  - Purpose.
  - Reader outcome.
  - Depth.
  - Tone.
  - Excluded topics.
  - Primary keyword.
  - Competitor angle.
  - Evidence standard.
  - Visual expectations.
- Use agent job path with fake runner fixtures for tests.
- Store editable questions and answers.
- Block plan generation until all required questions are answered and submitted.
- Frontend:
  - Boundary questions screen.
  - Partial save.
  - Required answer validation.

## Public Interface

- Editor generates questions for an article.
- Editor edits answers and submits them.
- Article moves to `BOUNDARY_ANSWERS_SUBMITTED`.

## Acceptance Criteria

- [ ] TDD pre-requirement: before implementation, write one failing API behavior test for required unanswered questions blocking submission; proceed through generate/save/submit behavior one cycle at a time and record evidence in `Result`.
- [ ] Questions are generated from article brief plus target site config.
- [ ] Required categories are represented.
- [ ] Editor can save partial answers.
- [ ] Required unanswered questions block submission.
- [ ] Submission writes workflow event and updates article status.
- [ ] Plan generation remains blocked until boundary answers are submitted.
- [ ] Frontend displays required/optional state clearly.

## Verification

- Run boundary question API tests.
- Run frontend page tests.
- Run a Docker Compose smoke flow from article creation to boundary answer submission.

## Result

- Status: Pending execution.
- TDD plan: To be filled during execution.
- Red evidence: To be filled during execution.
- Green evidence: To be filled during execution.
- Refactor notes: To be filled during execution.
- Verification output: To be filled during execution.