tpc-server

[x] ## **v1.0 - Basic Thought Logger** **Tests:** ```javascript // Test 1: Can POST a thought POST /thoughts Body: {"content": "Test thought"} Expect: 201 Created, {id: "1", content: "Test thought", timestamp: <current_time>} // Test 2: Thought is persisted GET /thoughts (even though not official yet, test internally) Expect: [{id: "1", content: "Test thought", timestamp: <time>}] // Test 3: Empty content fails POST /thoughts Body: {"content": ""} Expect: 400 Bad Request // Test 4: Missing content fails POST /thoughts Body: {} Expect: 400 Bad Request ``` [x] ## **v1.1 - Basic Plan Creator** **Tests:** ```javascript // Test 1: Can POST a plan POST /plans Body: {"title": "Test Plan", "description": "Do something"} Expect: 201 Created, {id: "1", title: "Test Plan", description: "Do something", status: "proposed", timestamp: <time>} // Test 2: Plan is persisted GET /plans (internal test) Expect: [{id: "1", title: "Test Plan", status: "proposed", ...}] // Test 3: Missing title fails POST /plans Body: {"description": "No title"} Expect: 400 Bad Request ``` [x] ## **v1.2 - Plan Status Updater** **Tests:** ```javascript // Test 1: Can update plan status PATCH /plans/1 Body: {"status": "in_progress"} Expect: 200 OK, {status: "in_progress"} // Test 2: Status change persists GET /plans/1 Expect: {status: "in_progress", ...} // Test 3: Invalid status fails PATCH /plans/1 Body: {"status": "invalid_status"} Expect: 400 Bad Request // Test 4: Non-existent plan returns 404 ``` [x] ## **v1.3 - Simple Retrieval** **Tests:** ```javascript [x] // Test 1: GET /thoughts returns all thoughts Create 3 thoughts, then: GET /thoughts Expect: 200 OK, array with 3 thoughts in chronological order [x] // Test 2: GET /plans returns all plans Create 2 plans, then: GET /plans Expect: 200 OK, array with 2 plans [x] // Test 3: Empty arrays when no data GET /thoughts (with empty DB) Expect: 200 OK, [] ``` [x] ## **v1.4 - Thought-Plan Linking** **Tests:** ```javascript // Test 1: POST /thoughts with plan_id (valid/invalid/missing) persists correctly POST /thoughts {"content": "Linked", "plan_id": "1"} -> 201 includes plan_id POST /thoughts {"content": "Unlinked"} -> 201 no plan_id POST /thoughts {"content": "Invalid", "plan_id": "999"} -> 201 includes plan_id "999" // Test 2: GET /plans/:id/thoughts returns linked thoughts sorted asc by timestamp Create plan 1, add 2 linked thoughts -> GET /plans/1/thoughts -> 200 array sorted asc, length 2 // Test 3: Filters unlinked thoughts, empty for no linked thoughts Add unlinked thought -> GET /plans/1/thoughts -> only linked ones // Test 4: Invalid/non-existent plan_id returns 200 with [] GET /plans/999/thoughts -> 200 [] // Test 5: Orphans (thoughts with invalid plan_id) not returned for that plan POST thought with plan_id "999" -> GET /plans/999/thoughts -> 200 [] // Test 6: Integration: Create plan, link thoughts, retrieve sorted // Test 7: Regression: Prior endpoints (POST/GET plans/thoughts) unaffected ``` [x] ## **v1.5 - Plan Changelog** **Tests:** ```javascript // Test 1: Can append to changelog PATCH /plans/1/changelog Body: {"change": "Added feature X"} Expect: 200 OK, {changelog: ["Added feature X"]} // Test 2: Changelog accumulates PATCH /plans/1/changelog Body: {"change": "Fixed bug Y"} Expect: {changelog: ["Added feature X", "Fixed bug Y"]} // Test 3: Changelog appears in GET /plans/1 GET /plans/1 Expect: {changelog: [array of changes], ...} ``` [x] ## **v1.6 - Context Window Simulation** **Tests:** ```javascript // Test 1: GET /context returns incomplete plans Create: plan1 (proposed), plan2 (completed), plan3 (in_progress) GET /context Expect: {plans: [plan1, plan3], thoughts: [...]} // Test 2: Returns last 10 thoughts Create 15 thoughts GET /context Expect: thoughts array length = 10, with most recent thoughts // Test 3: Empty state GET /context (empty DB) Expect: {plans: [], thoughts: []} ``` [x] ## **v1.7 - SQLite Migration** **Tests:** ```javascript // All previous v1.0-v1.6 tests should still pass identically // Test: Data persists after server restart // Test: Concurrent requests don't corrupt data ``` [x] ## **v1.8 - Basic Filtering** **Tests:** ```javascript [x] // Test 1: Filter plans by status Create: plan1 (proposed), plan2 (in_progress), plan3 (proposed) GET /plans?status=proposed Expect: [plan1, plan3] [x] // Test 2: Limit thoughts Create 25 thoughts GET /thoughts?limit=5 Expect: array length = 5, oldest first (asc timestamp) [x] // Test 3: Invalid parameters handled gracefully GET /plans?status=invalid Expect: all plans (ignores invalid status) ``` [x] ## **v1.9 - Timestamp Queries** **Tests:** ```javascript [x] // Test 1: Get plans since timestamp Create plan1, wait 1s, note timestamp, create plan2 GET /plans?since=<timestamp> Expect: [plan2] only (plans >= date, sorted asc) [x] // Test 2: Get thoughts since timestamp Similar test with thoughts (thoughts >= date, sorted asc) [x] // Test 3: Invalid timestamp format GET /plans?since=not-a-timestamp Expect: all plans (ignores invalid, returns all) ``` ## **v2.0 - Static HTML UI** **Tests:** [x] ```javascript // Test 1: HTML file exists and loads GET / (or /index.html) Expect: 200 OK, Content-Type: text/html // Test 2: HTML contains expected elements Inspect response body for: - <title> containing "TPC Server" - Elements with ids like "plans-list", "thoughts-list" // Test 3: Can load CSS/JS assets // Test 4: UI loads and displays data from DB via sql.js // Test 5: /tpc.db route serves binary DB file // Test 6: No error messages in UI ``` [x] ## **v2.1 - Read-Only API UI** **Tests:** ```javascript // Test 1: UI makes API calls Load UI in browser, check network requests to: GET /plans, GET /thoughts // Test 2: Data displays in UI Create test data, load UI, verify plans and thoughts appear // Test 3: UI handles empty state Load with empty DB, verify appropriate message ``` [x] ## **v2.2 - Plan Detail Pages** **Tests:** ```javascript // Test 1: Clicking plan shows detail view UI test: Click plan in list, verify detail page loads // Test 2: Detail page shows plan data Verify title, description, status, changelog all display // Test 3: Related thoughts appear Create thought linked to plan, verify it appears on detail page // Test 4: Back navigation works ``` ## **v2.3 - Plan Editing API** **Tests:** ```javascript // Test 1: Can update plan details PUT /plans/1 Body: {"title": "New title", "description": "New desc"} Expect: 200 OK, {title: "New title", ...} // Test 2: Partial updates work PUT /plans/1 Body: {"title": "Only title updated"} Expect: 200 OK, description unchanged // Test 3: Invalid data rejected PUT /plans/1 Body: {"title": ""} Expect: 400 Bad Request ``` ## **v2.4 - The "Dirty Flag" System** **Tests:** [x] ```javascript // Test 1: Schema backfill sets needs_review=0 for existing plans after migration // Verify migration UPDATE sets 0 where NULL // Test 2: PUT /plans/:id sets needs_review=1 (human), returns with flag PUT /plans/1 {"title": "Updated"} Expect: 200 {needs_review: 1, last_modified_by: "human", ...} // Test 3: Agent PATCH sets needs_review=0 PATCH /plans/1 {"status": "in_progress"} GET /plans/1 Expect: needs_review: 0, last_modified_by: "agent" // Test 4: GET /plans and /plans/:id include needs_review GET /plans Expect: array with "needs_review" property each // Test 5: Integration: Edit plan, verify flag in GET /plans?status=..., /context PUT /plans/1 (human edit) GET /plans?status=proposed -> find plan with needs_review: 1 GET /context -> incompletePlans include needs_review: 1 // Test 6: Filters compatible (e.g., ?status, ?since include flag) GET /plans?status=proposed&since=0 Expect: plans with needs_review ``` Full suite passes: npm test (167 tests) && npx playwright test (8 UI tests, unchanged). [x] ## **v2.5 - Agent Review System** **Tests:** ```javascript [x] // Test 1: Can filter by needs_review (unit tests for filtering) Create plans with different needs_review values GET /plans?needs_review=true Expect: only plans with needs_review=true [x] // Test 2: Agent can clear flag (unit tests for PATCH) PATCH /plans/1 Body: {"needs_review": false} GET /plans/1 Expect: {needs_review: false} [x] // Test 3: Context endpoint includes needs_review plans (unit tests for context) GET /context Expect: includes plans with needs_review=true in plans array ``` [x] ## **v2.6 - Rich Text Support** **Tests:** ```javascript // Test 1: Markdown in plan descriptions PUT /plans/1 Body: {"description": "**Bold** and *italic*"} GET /plans/1 Expect: description contains same markdown // Test 2: UI renders markdown as HTML Load plan detail, verify <strong>Bold</strong> appears // Test 3: Plain text still works Verify non-markdown content displays normally // Regression: All prior E2E tests (v2.0-v2.5) pass after modular refactoring of server.js npx playwright test → 11 passed (plans count: 10, thoughts count: 6 from migration) npm test → All Jest tests (v1.0-v2.6) pass, confirming unit/integration logic intact post-refactoring ``` [x] ## **v2.6 - Rich Text Support** **Tests:** ```javascript // Test 1: Markdown in plan descriptions PUT /plans/1 Body: {"description": "**Bold** and *italic*"} GET /plans/1 Expect: description contains same markdown // Test 2: UI renders markdown as HTML Load plan detail, verify <strong>Bold</strong> appears // Test 3: Plain text still works Verify non-markdown content displays normally ``` ## **v2.7 - Search & Organization** **Unit Tests (Jest in v2.7.test.js):** - **Schema Migration:** Verify idempotent addition of `tags` column to plans/thoughts (no error on re-run; existing rows backfilled to `[]`). - **Tagging Operations:** - POST /plans/:id/tags {tags: ["urgent"]} → 200, plan.tags = '["urgent"]'; invalid tags (e.g., uppercase/special chars) normalized or rejected. - PATCH /plans/:id/tags {action: "add", tags: ["high-priority"]} → adds without duplicates; {action: "remove", tags: ["urgent"]} → removes if exists. - Edge: Empty tags array clears all; non-existent ID → 404; invalid action → 400. - Similarly for thoughts. - **Search Logic (GET /search):** - q="React" on plan title/desc/thought content/tags → matches partial/case-insensitive (use LIKE or FTS5); relevance score (e.g., term count + tag exact match bonus). - Combined results: {plans: [matched plans sorted by score DESC then timestamp DESC], thoughts: [...], total: N}; ?type=plans → only plans. - ?limit=5 → truncates results; empty q → returns top 20 recent (timestamp DESC, no search). - No results → {plans: [], thoughts: [], total: 0}; ?tags=urgent → filters to tagged items only. - **Filtering (GET /plans?tags=):** - ?tags=urgent → plans where JSON tags contains "urgent"; ?tags=urgent,high → AND (both tags present). - Invalid tags ignored; combined with ?status, ?since. - Similarly for thoughts. - **Context Enhancement (GET /context?search=):** - ?search="React" → incomplete plans + last 10 thoughts filtered to matches; no param → full as before. - **Backward Compatibility:** All pre-v2.7 tests pass (e.g., POST plan without tags succeeds; GET /plans omits tags if not requested). - **Optimizations:** Verify indexes created in migration; query performance (mock DB, assert no full scans). **E2E Tests (Playwright in e2e/v2.7.test.js):** - Load UI, enter search "React" in input → verify /search request, results render in accordions (plans section shows matched plans with tag badges; thoughts section similar). - Test empty query → shows recent/all; no results → "No matches" message. - Tag filter: If UI added (e.g., dropdown), select "urgent" → filtered list; click plan detail → shows tags. - Integration: Create plan via API, add tag, search by tag → appears in UI; markdown in results renders correctly. - Cross-browser: Test rendering on Chrome/Firefox; no console errors. - Regression: Existing UI flows (plan list, details, thoughts) unchanged. Full suite: `npm test` (add ~20 new tests to 167+); `npx playwright test` (add 5-7 new UI tests to 8+). Ensure 100% pass post-v2.7. ## **v2.8 - The "Endgame"** **Tests:** ```javascript // Test 1: Real-time updates Open two UI windows, edit plan in one, verify other updates automatically // Test 2: Plan templates work POST /plans/templates/bugfix Expect: creates plan with pre-filled bugfix template // Test 3: Export functionality GET /export/plans.json Expect: 200 OK, JSON file with all plan data // Test 4: Error handling Test various error conditions, verify graceful degradation ``` ## **Testing Strategy Notes:** - **Each version's tests become regression tests** for subsequent versions - **API tests** can be automated with Jest/Supertest - **UI tests** can use Playwright/Cypress from v2.0 onward - **Integration tests** verify data flows between AI agent ↔ server ↔ human UI - **Performance tests** become important around v1.7 (SQLite migration)