# ChatStack

> Machine-readable home summary. HTML: https://www.chatstack.app/. Canonical reference: https://www.chatstack.app/llms-full.txt.

ChatStack is an interview-driven 4-agent workflow that turns a product idea into structured user stories, functional and non-functional requirements, technical specifications, and a story-point cost estimate. Outputs are JSON (canonical), plus Markdown and CSV, designed for Cursor, Claude Code, and the Model Context Protocol.

Prefer APIs and Markdown over scraping HTML.

## Start here

- Human chat: https://www.chatstack.app/
- Full agent reference: https://www.chatstack.app/llms-full.txt
- Discovery: https://www.chatstack.app/api/agent
- OpenAPI: https://www.chatstack.app/openapi.json
- Sitemap: https://www.chatstack.app/sitemap.xml

## Product pages

- What is ChatStack: https://www.chatstack.app/what-is-chatstack.md
- Popular Apps (MVP catalog + TCO calculator): https://www.chatstack.app/popular-apps.md
- Cost model: https://www.chatstack.app/how-much-does-it-cost-to-build-an-app.md
- Cursor / MCP integration: https://www.chatstack.app/cursor-ai-prd-integration-guide.md
- Services: https://www.chatstack.app/ai-development-services.md
- FAQ: https://www.chatstack.app/faq.md
- About: https://www.chatstack.app/about.md
- Speak to a human: https://www.chatstack.app/speak-to-a-human.md
- NDA and terms (HTML only): https://www.chatstack.app/non-disclosure-and-tcs

## Agent paths (short)

| Path | Auth |
|------|------|
| Public interview MCP | Guest token from `start_session` |
| Interview REST `/api/agent/interview/*` | Portal PAT or guest `interview_token` |
| Skip-interview `POST /api/agent/submit` | `AGENT_API_KEY` (never in MCP config) |
| Portal MCP / CRM REST | Portal personal access token |

Browser UI automation notes: https://www.chatstack.app/agent-browser.md