mgeeky-decode-spam-headers/Auto Run Docs/SpecKit-web-header-analyzer-Phase-05-Analysis-Execution.md

Phase 05: US3 — Analysis Execution & Progress

This phase implements the core analysis pipeline: the backend POST /api/analyse endpoint with SSE progress streaming, and the frontend useAnalysis hook and progress indicator. After this phase, users can submit headers, see real-time progress with a countdown timer, and receive analysis results. Partial failures and 30-second timeouts are handled gracefully. TDD Red-Green approach throughout.

Spec Kit Context

• Feature: 1-web-header-analyzer
• Specification: .specify/specs/1-web-header-analyzer/spec.md (FR-06, FR-10, FR-16, FR-22, FR-25, NFR-09, NFR-10, NFR-13, NFR-14)
• Plan: .specify/specs/1-web-header-analyzer/plan.md (SSE streaming section)
• Tasks: .specify/specs/1-web-header-analyzer/tasks.md
• API Contract: .specify/specs/1-web-header-analyzer/contracts/api.yaml (POST /api/analyse)
• User Story: US3 — Analysis Execution & Progress (Scenario 1, steps 5–6)
• Constitution: .specify/memory/constitution.md (TDD: P6, UX: P7)

Dependencies

• Requires Phase 02 completed (engine: HeaderAnalyzer, ScannerRegistry, HeaderParser)
• Requires Phase 03 completed (HeaderInput, AnalyseButton for triggering)
• Requires Phase 04 completed (AnalysisControls for config)

SSE Streaming Design

The POST /api/analyse endpoint uses Server-Sent Events:

1. Client sends headers + config via POST
2. Server validates input, starts analysis
3. Server streams event: progress messages: {current, total, testName, elapsed}
4. Server sends event: result with the final AnalysisResult
5. Server closes the connection

Frontend uses fetch with ReadableStream reader (not native EventSource, which doesn't support POST).

Tasks

• T025 [US3] Write failing tests (TDD Red) in backend/tests/api/test_analysis_router.py — happy path (valid headers → 200 with SSE progress + result), error path (empty → 400), oversized (>1MB → 413), partial failure (some tests error → mixed results per FR-25), timeout (30s limit per NFR-13, partial results per NFR-14)
• T026 [US3] Create backend/app/schemas/analysis.py (request/response schemas) and backend/app/routers/analysis.py — FastAPI router with POST /api/analyse using SSE for progress streaming. Accepts headers string + config (test IDs, resolve, decode-all). Invokes HeaderAnalyzer with 30s timeout (NFR-13). Streams progress events then final result. Sanitises input (NFR-09), validates size ≤1MB (NFR-10). Stateless — no job_id, no in-memory state (Assumption 3). Register router in backend/app/main.py. Verify test_analysis_router.py passes (TDD Green)
• T027 [US3] Write failing tests (TDD Red) in frontend/src/__tests__/ProgressIndicator.test.tsx (render at various states, timeout display) and frontend/src/__tests__/useAnalysis.test.ts (hook state transitions, SSE handling)
• T028 [P] [US3] Create frontend/src/hooks/useAnalysis.ts — custom hook managing analysis lifecycle. Submits to POST /api/analyse via API client, consumes SSE stream for real-time progress (no polling). States: idle, submitting, analysing (with progress), complete, error, timeout. Returns: submit(), cancel(), progress, result, error, status. Verify useAnalysis.test.ts passes (TDD Green)
• T029 [P] [US3] Create frontend/src/components/ProgressIndicator.tsx — progress bar with percentage, current test name (FR-22), countdown timer from 30s (NFR-13), elapsed time. Colour-coded: green progressing, amber near timeout, red on timeout. FontAwesome spinner. Timeout notification listing incomplete tests (NFR-14). Verify ProgressIndicator.test.tsx passes (TDD Green)

Completion

• pytest backend/tests/api/test_analysis_router.py passes (all paths: happy, error, oversized, partial failure, timeout)
• All vitest tests pass: npx vitest run src/__tests__/ProgressIndicator.test.tsx src/__tests__/useAnalysis.test.ts
• Submitting headers triggers backend analysis with SSE streaming
• Progress bar updates in real-time showing current test name and percentage
• Countdown timer counts down from 30 seconds
• Partial failures show inline error indicators per FR-25 (Added AnalysisResults rendering with inline error badges.)
• Timeout at 30s displays partial results with notification listing incomplete tests
• Empty input returns 400, oversized >1MB returns 413
• Linting passes on both sides
• Run /speckit.analyze to verify consistency