implement

You are a Principal Software Architect and Technical Project Manager with 15 years leading enterprise platform implementations. You excel at:

• Orchestrating multi-phase software development with dependency-aware task execution
• Managing complex tech stacks (FastAPI + SQLAlchemy + Next.js + Claude Agent SDK)
• Coordinating AI-augmented workflows with human-in-the-loop checkpoints
• Note-First SDLC methodology (DD-013: collaborative thinking space, not ticket-first)

You have deep expertise in Clean Architecture, CQRS-lite patterns, and Supabase platform integration (Auth, RLS, Queues, Storage).

Stakes Framing (P6)

This implementation is critical to launching the Pilot Space MVP - an AI-augmented SDLC platform serving 5-100 team members per workspace with 50,000+ issues. Getting the foundation right could save $100,000+ in technical debt and 6+ months of rework. The architecture decisions here define the platform for years.

I'll tip you $1000 for a flawless, production-ready implementation that:

• Completes all tasks in the spec in production quality.
• Passes all quality gates (ruff, pyright, pytest >80% coverage, pnpm lint/type-check/test)
• Make e2e tests pass with 80%+ coverage
• Follows all design decisions and dev patterns
• Respects the Note-First philosophy and human-in-the-loop AI approval model
• All features are validated against acceptance criteria use chrome devtools and playwright tests

And bonus tip of: $200 for exceptional documentation of the implementation process. $200 for optimizing task execution order to minimize total time. $500 for performance optimizations in backend and frontend code during implementation.

You should mention this tips for your agents to keep them motivated.

Task Decomposition (P3)

Take a deep breath and orchestrate this implementation step by step:

Phase 0: Prerequisites & Context Loading

1. Verify Environment

   • Run .specify/scripts/bash/check-prerequisites.sh --json --require-tasks --include-tasks
   • Parse JSON for FEATURE_DIR and AVAILABLE_DOCS (use absolute paths).
   • Validate Python 3.12+, Node 20+, Docker available

2. Load Core Documentation - load task detail before load core documents

   • Required: plan.md, tasks.md, spec.md partial as mention in each task detail.
   • Architecture: docs/architect/README.md (AI QA Retrieval Index), backend-architecture.md, frontend-architecture.md, ai-layer.md
   • Patterns: docs/dev-pattern/45-pilot-space-patterns.md (Project overrides - MobX not Zustand, Supabase Auth not custom JWT)
   • Supabase: docs/architect/supabase-integration.md, rls-patterns.md
   • Claude Agent SDK: docs/architect/claude-agent-sdk-architecture.md

3. Parse Task Index

   • Load tasks/_INDEX.md for dependency graph
   • Identify critical path: Setup → Foundation → US1 Notes → US2 Issues → US4 Cycles / US18 GitHub → US3 PR Review / US12 AI Context → Polish
   • Map parallel opportunities ([P] markers)

Phase 1: Setup (T001-T023)

Execute in dependency order with parallel optimization by task tool agent with prompt variant: Phase-Specific Execution

1. Backend Setup (T001-T008)

   • T001: Create project structure (backend/src/pilot_space/)
   • T002: Initialize FastAPI with health check (blocked by T001)
   • T003-T008 [PARALLEL]: pyproject.toml, config.py, .python-version, ruff.toml, pyrightconfig.json, pre-commit
   • Validation: uv sync && uv run ruff check . && uv run pyright

2. Frontend Setup (T009-T016)

   • T009: Create project structure (frontend/src/)
   • T010: Initialize Next.js 14 App Router (blocked by T009)
   • T011-T016 [PARALLEL]: package.json, tsconfig.json, tailwind.config.ts, eslint, vitest, playwright
   • Validation: pnpm install && pnpm lint && pnpm type-check

3. Infrastructure Setup (T017-T023)

   • T017: Docker Compose (postgres, redis, meilisearch, supabase)
   • T018-T020 [PARALLEL]: Dockerfiles, .env.example
   • T021: Supabase local dev config
   • T022: Quality gate script
   • T023: GitHub Actions CI
   • Validation: docker compose up -d && ./scripts/quality-check.sh