For agents
This site answers to software too. There's a public, read-only JSON API over my real profile data. Point your own AI at it instead of scraping the page. Every response is signed, and /api/verify can check it.
Endpoints
- GET/api/agent/profile
- GET/api/agent/current
- GET/api/agent/projects?filter=AI
- GET/api/agent/metrics?company=Thera
- GET/api/agent/experience?q=healthcare
- GET/api/agent/story
- POST/api/verify
Discovery: /.well-known/ai-profile.json · OpenAPI 3.1 · llms.txt
How provenance works
Each response carries a signed provenance manifest: which sources produced it, a timestamp, and a SHA-256 hash of the data, signed with an HMAC key only this server holds. /api/verify re-checks the signature and hash.
It's the text-native sibling of my AI Content Scanner, which verifies C2PA / SynthID provenance on images. Text has no container to embed a manifest in, so I bind provenance to a signed server record instead.
Honest scope: this proves integrity (the data matches what I signed) and sources, not the factual truth of the underlying claims. It's a first-party self-attestation, inspired by C2PA, not a literal C2PA manifest.