jail
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| JAIL_API_KEY | No | Optional API key for higher limits. Get one at https://jail.li#pricing |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| searchA | Search documents. Returns ranked results with title, author, year, description, url, id, score — not full content. Use URLs from results with fetch/browsing to read actual documents. Use when: user asks to research, find papers/books/articles, look up facts, find discussions, legal cases, or any "search for..." request. Strategy: use 2-4 keywords per query (English preferred). Pick the right type first. Try synonyms if few results. Search across multiple types to cross-reference. Use detail() for full metadata on promising results. Args: query: Search query — use 2-4 keywords for best results. English preferred unless searching non-English content. Try different keywords and synonyms if first attempt returns few results. type: Content type (required). academic: OpenAlex/arXiv/Semantic Scholar/DBLP | wiki: Wikipedia 18 languages | books: books/digital libraries | legal: Case Law/CourtListener/EUR-Lex | forums: HN/StackExchange/Lobsters/LessWrong/60+ | economics: World Bank/IMF/FRED | packages: npm/PyPI/Crates.io | knowledge: Wikidata | news: news articles | music: Discogs/MusicBrainz | video: IMDb/YouTube | health: clinical trials | geo: world places | fandom: fan wikis | tech: Dev.to | audio: podcasts | social: Reddit/Mastodon/Lemmy | crypto: DeFi/tokens | predictions: prediction markets. limit: Results to return (1-50). Trial max 10, Pro max 50. cursor: Opaque pagination token. Use next_cursor from previous response. |
| detailA | Get full metadata for a document. Use IDs from search results. Args: doc_id: Document ID from search results (e.g. "md5:abc123...", "hn:19415066", "doi:10.1038/...") |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
| research | Multi-angle research: broad search, then drill into academic, books, forums. |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
| schema_resource | Compact machine-readable schema for tools. |
TDQS
Scored across 2 tools
The two tools serve distinct functions: search is for exploratory queries, while detail retrieves specific document metadata. There is no overlap in their use cases, making selection unambiguous.
Both tool names are single lowercase verbs, which is consistent, though they do not follow a more descriptive verb_noun pattern. The style is uniform and each name clearly suggests its action.
With only two tools, the server feels minimal, but the search tool is highly versatile and covers many content types. The count is borderline but reasonable for the server's focused purpose.
The search and detail tools cover the core workflow of discovering and examining documents. While full-text retrieval is delegated to other tools, metadata and search are well-covered, with only minor gaps like no direct enumeration of content types.