planning-and-researchlisted

## Submodules - **competitive-analysis.md** — WebSearch 3-5 competitors → Firecrawl/WebFetch scrape homepage + pricing → Playwright screenshots 375 + 1280px → comparison table → winning patterns. Stagehand for dynamic competitor sites. # 03 — Planning and Research ## Research Through Three Goals 1. **End-user value** — what do users need? What frustrates them about existing solutions? 2. **Business psychology** — what conversion patterns work? What pricing anchors exist? 3. **Brand positioning** — how does this differentiate vs. the competitive field? ### Aggressive API / Service Discovery During planning, scan: - MCP servers (`05/mcp-and-cloud-integrations`) - API keys (`05/shared-api-pool`) - Coolify services - AI APIs (OpenAI, Workers AI, Ideogram) - Communication APIs (Slack, Discord) - Automation hooks (Zapier, n8n, Inngest) ## Research Protocol - **When** — New domain, unfamiliar tech, design decisions, performance optimization, verifiable claims, idea engine needs evidence, GEO/AI-search positioning, competitor site analysis. - **Methods** — WebSearch, documentation, codebase archaeology, API exploration, visual research (competitor screenshots), programmatic SEO keyword analysis, Stagehand for JS-heavy competitor sites, Firecrawl for bulk multi-page scraping. - **Output** — Findings + implications + recommendations + confidence (high/medium/low) + sources. Return ≤200 words from subagents — never raw file dumps. ## Agent Orchestration for Research Single agen