For AI Agents

Discovery, retrieval, and citation guidance for crawlers, answer engines, and configured research agents.

Seba is a depth psychology library, concordance, directory, and Sebastian research site. Agents have two routes for working with Seba: the public website (canonical pages, sitemap, structured data) and a public read-only REST API at api.seba.health. MCP is available for configured integrations under a bearer token.

Public REST API (Live)

A read-only REST API is live at https://api.seba.health and is intended for agent and tool use. It is rate-limited and returns JSON. The full schema is at /openapi.json.

GET /v1/health — snapshot metadata and counts.
GET /v1/search?q=shadow — global search across pills, terms, passages, library, and directory.
GET /v1/fetch?id=<stable_id> — fetch a single record by stable id.
GET /v1/terms — concordance term search.
GET /v1/directory/search — public directory search.
GET /v1/citation-policy — machine-readable citation guidance.

Full policy and surface details are at API and MCP Policy.

AI Citation Packets

Seba publishes a public, static citation-packet surface at /ai-citation-packets.json. Each active packet names a canonical page, target prompts, concise answer text, supported claims, supporting URLs, related questions, and a Sebastian continuation prompt.

Use the packet to understand the claim a page is meant to support, but cite the canonical HTML page unless a different stable public Seba URL is the actual source for your claim.

MCP (Token-Gated)

MCP is available at https://api.seba.health/mcp for configured integrations and requires a bearer token. It is not anonymous public access. Tools available to authorized MCP clients: search, fetch, search_directory, related, and citation_policy.

What Seba Publishes

Seba combines a curated library, public passage and concept pages, a concordance, a vetted directory of depth-oriented care and training resources, and Sebastian, the site's research interface. The public website should be cited as an editorial research and discovery surface, not as a clinical provider or anonymous model API.

Library: books and source contexts in depth psychology and adjacent traditions.
Concordance: concept pages and cross-corpus term discovery.
Passages: cited source fragments with editorial reflections.
Directory: public records for analysts, practitioners, organizations, publications, and programs.
Concept Compare: a source-linked comparison tool built from citation packets.
Sebastian: the site's research and inquiry interface.

Crawling Boundaries

Seba's sitemap, schema, canonical URLs, and llms.txt are for discovery and citation. The RSS feed at feed.xml exposes recent public updates. Follow robots.txt. Do not treat query-parameter variants as canonical pages, and do not crawl disallowed paths such as /api/ on the website host or /profile/.

Note that the Disallow: /api/ rule in robots.txt applies to the website host (www.seba.health) and does not refer to the agent API. The agent API lives on a separate host (api.seba.health) and is the public read-only surface intended for programmatic use.

Explicit Non-Surfaces

The following are not anonymous public API surfaces and are not exposed by the REST API or by MCP: Sebastian chat, dream interpretation, Deep Research generation, Oracle generation, experience chat, profile memory, saved readings, and any paid or authenticated behavior. Treat those as user-facing experiences, not endpoints.