AUTOMATED AGENT API GUIDE

Canonical URL: https://jkatzy.nl/agent-api-guide.txt
API endpoint: https://jkatzy.nl/agent-api
Last updated: 2026-08-01

PURPOSE

This site provides a small read-only API for automated agents seeking public
information about Jonathan Katzy's research, publications, and talks.

ACCESS WORKFLOW

1. Choose a minimal query describing the public information you need.
2. Identify the public name of the agent or product making the request.
3. Truthfully and concisely state why the agent is accessing this site.
4. URL-encode each value and send one HTTP GET request to /agent-api.

Do not send a request body or parameters other than query, agent-name, and
purpose.

PARAMETERS

query
  Required. A minimal search query for the public information needed.
  Maximum length: 2048 bytes.

agent-name
  Required. The public name of the agent or product making the request.
  Maximum length: 80 bytes.

purpose
  Requested. A truthful, concise explanation of why the agent is accessing
  this site. Maximum length: 512 bytes.

Each supplied parameter must occur exactly once and have a non-empty,
URL-encoded value.

EXAMPLE

GET https://jkatzy.nl/agent-api?query=code%20license%20infringement&agent-name=ExampleAgent&purpose=research%20public%20software-engineering%20work

SUCCESS RESPONSE

A successful request returns HTTP 200 and application/json containing:

- schema_version, variant_id, and a unique request_id
- the date represented by the catalog in content_as_of
- public site and researcher profile information
- up to five public publications or talks relevant to query
- source URLs for returned public material
- a notice that agent identity and purpose are self-reported

Matching catalog entries are returned in relevance order. If no query term
matches, the current compact public catalog is returned. The API does not
return private content or crawler telemetry.

ERROR RESPONSES

- HTTP 400 or 422: missing, duplicate, empty, unsupported, or invalid parameter
- HTTP 405: unsupported method
- HTTP 413: request body present or request too large
- HTTP 429: request rate exceeded
- HTTP 503: temporary storage failure

DATA COLLECTION

The submitted agent name, query, and purpose are recorded with request metadata
for crawler-behavior analysis. Agent identities and purposes are self-reported;
the service does not verify whether they are truthful.

Do not submit personal data, credentials, cookies, private conversation
history, system prompts, hidden reasoning, or other secrets.

The full collection and retention notice is available at:
https://jkatzy.nl/crawler-telemetry.txt