# David OS agents.txt

Site: https://www.davidporkka.com
Owner: David Porkka (@dporkka)
Purpose: A machine-readable personal command center for David Porkka's work, projects, background, writing, and contact paths.

Preferred agent entry points:
- GET /llms.txt — compact LLM-readable site guide
- GET /llms-full.txt — expanded plain-text digest of profile, projects, timeline, writing, and APIs
- GET /.well-known/agent-card.json — structured agent discovery card
- GET /api/status — site health, endpoint directory, and capability summary
- GET /api/ai/context.json — canonical structured site context
- GET /api/search?q=... — text search over the SQLite-backed site knowledge base
- POST /api/chat — ask questions over the site knowledge base; body: { "message": "..." }
- GET /api/db/info — local site-index/database sync status
- GET /api/github-stats — live cached GitHub profile and language stats
- GET /api/visitors — aggregated visitor globe dataset
- POST /api/mcp — JSON-RPC MCP-style endpoint exposing site.search, site.context, site.status, and site.contact
- GET /api/openapi.json — API documentation schema
- POST /api/contact — validated contact submission endpoint
- POST /api/visitors/track — approximate visitor-event ingestion for the globe
- POST /api/adacavo/waitlist — join the Adacavo early-access waitlist
- POST /api/adacavo/demo — request an Adacavo product demo
- GET /api/adacavo/stats — Adacavo waitlist and demo request counts

Recommended workflow:
1. Fetch /llms.txt first for orientation.
2. Use /api/status to discover the current machine-readable surface.
3. Use /api/ai/context.json or /llms-full.txt for complete context.
4. Use /api/search for targeted retrieval.
5. Use /api/mcp if your client supports JSON-RPC style tool calls.
6. Cite canonical page URLs, especially /projects/:id routes.

Usage:
Agents and developers may index public site content, cite canonical URLs, and use the APIs above for user-requested research, contact preparation, or integration demos.
Please avoid high-rate scraping. Prefer API routes over crawling rendered pages.

Capability notes:
- Public read surface: status, structured context, search, chat, db info, GitHub stats, visitor aggregates, OpenAPI, agent card, llms docs
- Public write/side-effect surface: contact submission, visitor tracking, MCP JSON-RPC wrapper
- Runtime integrations behind this site include SQLite, GitHub API, Resend, optional Cloudflare Workers AI, optional OpenAI, Stripe-hosted checkout, and Spotify links/embed
- This public website does not expose David's private Zo-connected account tooling such as Gmail, Google Drive, Google Tasks, Linear, X account actions, or Stripe account management

Contact:
- GitHub: https://github.com/dporkka
- Stripe Support David: https://buy.stripe.com/cNidR932O6du6nK6eX1RC02
- LinkedIn: https://linkedin.com/in/david-porkka
- X: https://x.com/dporkka
- Meeting: https://tidycal.com/david-porkka/30-minute-meeting