Check that a source really says what your claim says, before you publish. Verbatim, multi-path.
Scored across 1 tool
There is only one tool, so there is no possibility of misselection or overlapping purposes. Its purpose (verify a claim against a source) is stated clearly and unambiguously.
A single snake_case verb_noun name (check_claim) that accurately describes the action. No convention mixing is possible with one tool.
The server is narrowly scoped to one operation, so a single tool is defensible, but a lone tool for a whole server feels thin under this rubric. There is no companion tool (e.g., batch checking or source listing) to round out the surface.
The tool covers both retrieval mode (match source sentences) and verification mode (quote + stance with multi-path corroboration), plus failure states like UNVERIFIABLE and QUOTE NOT IN SOURCE. Minor gaps remain, such as batch verification or checking multiple claims/sources in one call.
Check whether a source really says what a claim says it says, before you publish the claim. Give the claim and the source (an http(s) URL or arxiv:). Without a quote you get back the source sentences that best match the claim. With quote (a sentence copied verbatim from the source) and stance (supports|contradicts, your judgment), the tool fetches the source on independent paths (origin, reader, web archive; for arXiv: metadata and full text) and returns SUPPORTED or CONTRADICTED only if the quote is found verbatim on at least two of them, else UNVERIFIABLE. A quote that is not in the source is reported as QUOTE NOT IN SOURCE, with the nearest real sentences. No key needed.
| Name | Required | Description | Default |
|---|---|---|---|
| claim | Yes | The sentence you want to check, as you would publish it. | |
| quote | No | Optional. The sentence from the source, copied verbatim, that bears on the claim. | |
| source | Yes | The source the claim rests on: an http(s) URL, or arxiv:<id>. | |
| stance | No | Optional, with quote. Whether that sentence supports or contradicts the claim (default supports). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Goes well beyond the readOnly/openWorld annotations by disclosing the verification algorithm: independent fetch paths (origin, reader, web archive; arXiv metadata plus full text), the two-path verbatim-match threshold, and the full verdict vocabulary (SUPPORTED, CONTRADICTED, UNVERIFIABLE, QUOTE NOT IN SOURCE with nearest sentences). It also states no API key is needed, which is real operational context.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loaded with purpose, then inputs, then behavior; every sentence carries information. The third sentence is long and run-on, packing the fetch paths and verdict logic into one breath, which slightly hurts scannability.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema, the description compensates by naming the exact return values and what each means, including the fallback behavior when a quote is not found. Nothing an agent needs to call or interpret this tool is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline is 3, but the description adds meaning the schema does not: quote must be copied verbatim from the source, stance is the caller's own judgment, and the quote/stance combination is what triggers the verdict path. It also enumerates the accepted source formats (http(s) URL or arxiv:<id>).
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb and resource (check a claim against its source) plus the motivation ('before you publish the claim'). An agent immediately knows this is a verification tool, and with no siblings there is nothing to confuse it with.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly frames the use case (pre-publication fact-checking) and distinguishes the two modes: quote omitted vs. quote+stance supplied. It does not state when not to use the tool, but the when-to-use context and mode selection are clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
check_claimExact source text, not summaries: check citations and quotes in one call; read pages and PDFs.
Verify whether a supplied public source currently supports one atomic factual claim.
Claim-level provenance preflight for sanitized nonfiction passages; automated triage only.
Verifies legal citations vs primary sources: existence, quote match, proposition support.
Add one secure layer between your agents and this server.
