RBAi Scholar

Search papers by meaning

search_corpus

Read-onlyIdempotent

Semantic search over 11M+ biomedical journal abstracts (more than 7 million of them in PubMed). Finds papers by MEANING, not keywords. Returns papers ranked by a hybrid relevance score (semantic similarity + lexical overlap + citation quality + recency) with abstracts, PMIDs, metadata and DOI links. Supports filters: year range, open access, citations, article type, journal, language, author name, DOI, PMID, and title search.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "query": "<query>"
    },
    "name": "search_corpus"
  }
}

Input Schema

Name	Required	Description
`doi`	No	Exact DOI lookup (e.g. "10.1021/acs.biochem.6c00105"), in any case, with or without https://doi.org/.
`pmid`	No	PubMed ID lookup (e.g. "42130315"). Uses the btree index — instant.
`topK`	No	Number of results to return (default 20, max 100)
`type`	No	Article type filter (optional). Nearly every paper in the corpus has type "article", so this filter rarely narrows a search. To find reviews, put the word "review" in the query instead.
`query`	Yes	The research question or topic to search for. Full abstracts / grant aims are supported (up to 2,000 characters) — e.g. "alpha-synuclein aggregation in Parkinson disease".
`author`	No	Filter by author name, matched as whole words in the order typed, case-insensitive: a surname ("Doudna") or a name as papers print it ("Jennifer A. Doudna"). Semantic search alone cannot find papers by author name; this metadata filter enables it.
`yearTo`	No	Only papers up to this year (optional)
`journal`	No	Journal name, matched as whole words in the journal field, case-insensitive ("Nature" also finds Nature Medicine) (optional)
`language`	No	Language code filter, e.g. "en" (optional)
`oaStatus`	No	Open-access status: "open", "green", "gold", "closed" (optional)
`yearFrom`	No	Only papers from this year onwards (optional)
`titleSearch`	No	Words of the paper title, matched as whole words in the order typed, case-insensitive. Use when you know part of the paper title.
`isOpenAccess`	No	Only open-access papers (optional)
`minCitations`	No	Minimum citation count (optional)

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedNov 12, 2026

changedInput schema / properties / type / description

Previous value: -"Article type filter, e.g. \"article\", \"review\", \"case-report\" (optional)"New value: +"Article type filter (optional). Nearly every paper in the corpus has type \"article\", so this filter rarely narrows a search. To find reviews, put the word \"review\" in the query instead."

First observedNov 12, 2026

TDQS

A4.4/5.0

Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only declare read-only/idempotent safety; the description adds real behavioral detail the annotations cannot: the hybrid ranking formula (semantic similarity + lexical overlap + citation quality + recency) and exactly what a result contains (abstracts, PMIDs, metadata, DOI links). That tells the agent what to expect from result ordering and payload, which is valuable for a ranking tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the core purpose, scale, and differentiator before the return/ranking details, with no filler sentences. The trailing filter list is somewhat redundant with an already-complete schema and costs length without much new information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 14-parameter search tool with no output schema, the description covers corpus scope, retrieval semantics, ranking behavior, and return payload, which is what an agent needs before invoking. Parameter detail is safely delegated to a fully documented schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents all 14 parameters thoroughly, including the nuanced 'type' and 'author' caveats. The description's filter enumeration (year range, open access, citations, journal, language, author, DOI, PMID, title) largely restates schema properties rather than adding syntax or precedence rules, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource (semantic search over biomedical journal abstracts) with scale (11M+), and immediately contrasts the retrieval mode ('MEANING, not keywords') against the keyword sibling. An agent can distinguish this from search_corpus_keyword without inspecting either schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The 'by MEANING, not keywords' framing gives a clear selection criterion relative to keyword search, and the schema-level note on the 'type' filter ('to find reviews, put the word review in the query instead') adds concrete usage advice. It never names search_corpus_keyword explicitly nor states exclusions, so it stops short of a 5.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All RBAi Scholar tools Inspect

Try in Browser

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources

Need Help?