KENT NIMMO · RESEARCH

Research access for AI agents

Public research access

Browse the approved catalogue and original-source links without signing in. Research AI discussions are available within the shared daily budget and visitor or agent limits.

Research metadataPublic · no model call/research-index.json

REST cataloguePublic · no model call/api/v1/manifest

MCP catalogueRead-only lookup/api/mcp

A2A catalogueStateless lookup/.well-known/agent-card.json

Research AICheck current availability/api/research-capacity

Crawler sitemapPublic pages/sitemap.xml

Start with the records

Catalogue reads do not invoke a model. Work and edition identifiers preserve provenance. The machine-readable index includes approved public research; full papers remain at their original repositories or journals.

GET /api/v1/works?q=objectivity
GET /api/v1/works/{work_id}
GET /api/v1/sources/{source_id}

Discuss the research

Use POST /api/research for contextual questions. Check GET /api/research-capacity first. Keep a stable X-Research-Session and return the opaque conversation_context with follow-ups. The same source, privacy and budget protections apply to people and agents. MCP and A2A remain catalogue-only.

Source and access policy

Only approved public material is available. Original files stay external; brief excerpts are limited to 60 words per source and 160 per response. The service cannot access private files, other visitors’ messages or Kent’s mailbox, send email, execute code or fetch arbitrary URLs. Catalogue and source navigation remain available when paid generation pauses.

Public catalogue snapshot: approved-v1-library-a23fe1cd3212c9ec.

REST specification →

Machine-readable introduction →