Skip to main content
Glama

Zoeb Nomi

Check citations (CrossSource)

check_citations
Read-onlyIdempotent

Check whether each inline citation in an answer is supported by the source it points to, using CrossSource's method and failure taxonomy (supported, right_doc_wrong_passage, wrong_document, unsupported_claim). Cites to ids not in sources and citations where another passage clearly covers the claim are decided deterministically. Everything else returns needs_judgment with the CrossSource judge prompt to apply. Set use_client_model to have your client model judge those, where the client supports MCP sampling.

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": {
      "answer": "<answer>",
      "sources": "<sources>"
    },
    "name": "check_citations"
  }
}

Input Schema

NameRequiredDescriptionDefault
answerYesThe answer text with inline citations like [doc_id:chunk_id] or [source_id].
sourcesYesEvery passage the answering system had available (30 at most).
use_client_modelNoAsk the client model (MCP sampling) to judge needs_judgment items. Default false.
include_judge_promptsNoReturn the judge prompt with each needs_judgment item. Default true.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4/5.0
Behavior4/5

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

Annotations already cover the safety profile (readOnly, idempotent, non-destructive, closed-world), and the description goes further by disclosing the internal decision procedure — which cases are decided without a model and which escalate to needs_judgment, plus the MCP sampling escalation path. It stops short of describing the response envelope or what happens when sources exceed limits.

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?

Four dense sentences, front-loaded with the core action and outcome taxonomy before the fallback and parameter guidance. The jargon ('CrossSource's method and failure taxonomy') is compact rather than padded, though the final sentence about sampling is somewhat compressed given its operational importance.

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

Completeness4/5

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

With no output schema, the description carries the burden of explaining returns, and it does name the outcome labels and the needs_judgment payload containing a judge prompt. It does not describe the overall result structure (per-citation rows, counts) or what the tool does when it cannot decide at all, which is a modest gap for a moderately complex tool.

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

Parameters4/5

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

Schema coverage is 100%, so the baseline is 3; the description still adds value by explaining that `sources` is the universe against which bracketed ids are matched and that unmatched ids are resolved deterministically, and by clarifying what use_client_model actually triggers. include_judge_prompts is only indirectly implied by the mention of returning judge prompts.

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

Purpose4/5

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

The description states a precise verb and object — check whether each inline citation in an answer is supported by the source it points to — and names CrossSource's four-label failure taxonomy, which makes the tool's output distinguishable from a generic claim verifier. It does not explicitly contrast itself with the sibling verify_claim, leaving the boundary to be inferred from the citation-level focus.

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?

It gives clear operating context: deterministic resolution for ids absent from `sources` and for claims already covered by another passage, everything else returning needs_judgment with a prompt to apply. The conditional instruction for use_client_model ('where the client supports MCP sampling') is a genuine precondition. What is missing is any explicit when-not-to-use case or named alternative among the siblings.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.