# henhao.ai · What's Actually Good in Biology

> henhao.ai indexes 363,000+ biology preprints and peer-reviewed papers across all 26 biology disciplines, enriched with structured experimental/computational tags, semantic embeddings, verified full-text digests, community ratings, and a reproducibility ledger.

## Agent & LLM Query Policy (Bounded Query Only, No Bulk Scraping)
- Interactive AI agents and research assistants MAY query our bounded read-only API and MCP server to answer user questions, verify citations, and discover papers.
- Every agent response using henhao data SHOULD include the canonical link (`https://henhao.ai/paper/<id>`).
- Bulk downloading, dataset mirroring, automated crawling of HTML pages, or unthrottled scraping is strictly prohibited. All query endpoints are rate-limited (30 req/min) and capped at <= 15 results per request.

## Model Context Protocol (MCP) Server
- **Endpoint (Streamable HTTP JSON-RPC 2.0):** `POST https://henhao.ai/api/mcp`
- **Discovery / Info:** `GET https://henhao.ai/api/mcp`
- **Available MCP Tools:**
  1. `search_biology_papers`: Search papers by keyword, structured tag, biology discipline, or DOI (max 15 results).
  2. `get_paper_details`: Fetch a single paper by henhao UUID or DOI (`10.1101/...`), including abstract, structured tags, reproducibility status, and related papers.
  3. `list_biology_taxonomy`: List the 26 biology disciplines and top structured experimental/computational tags.

## Bounded Read-Only JSON Endpoints
- `GET https://henhao.ai/api/v1/search?q=<query>&category=<discipline>&status=<all|journal|preprint>&limit=10`
- `GET https://henhao.ai/api/v1/papers/<id-or-doi>`
- `GET https://henhao.ai/doi/<doi>` (307 redirect to canonical paper page)
- `GET https://henhao.ai/rss.xml?category=<discipline>&status=<all|journal|preprint>`