Citation Intelligence MCP

citations_evidence

Read-onlyIdempotent

Extract the cited snippet from an AI engine's raw answer to see why each URL was referenced, returning surrounding context or the nearest quoted sentence.

Instructions

Extract the cited snippet from the AI engine's raw answer for each citation. Calls check_citations, then for each returned URL finds the first mention in raw_answer and returns a context window plus the nearest quoted span or containing sentence. Use to see why an engine cited a URL, not just that it did. Returns 'not found' for engines without raw_answer (Bing, Brave).

Input Schema

TableJSON Schema

Name	Required	Description	Default
`query`	Yes	Search query whose AI answer to extract citation evidence from.
`engine`	No	AI engine to query. web_rank engines (bing_serp, brave_serp) lack raw_answer and return no evidence.	auto
`max_results`	No	Max citations to extract evidence for.
`context_chars`	No	Half-width of the snippet window around each citation mention (chars). Total snippet is up to 2x this.

Output Schema

TableJSON Schema

Name	Required	Description
`note`	No
`query`	Yes	The query whose answer was analyzed.
`engine`	Yes	Engine used.
`evidence`	Yes	Per-citation evidence extracted from the raw answer.
`fetched_at`	Yes	UTC ISO-8601 timestamp.
`evidence_found`	Yes	Citations whose URL was located in the raw answer.
`has_raw_answer`	Yes	Whether the engine returned a raw answer.
`citations_total`	Yes
`raw_answer_chars`	Yes	Length of the engine's raw answer.

Schema Changelog

Changes observed during successful MCP inspections.

Addedv0.1.2Oct 2, 2026

TDQS

A4.2/5.0

Behavior4/5

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

Annotations already cover safety (readOnly, idempotent, openWorld). The description adds valuable behavioral detail beyond annotations: the internal call to check_citations, the window/quote extraction logic, and the engine-specific 'not found' behavior for Bing/Brave. Doesn't mention rate limits or performance, but adds meaningful context.

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?

Three dense sentences covering purpose, mechanism, and limitations. Front-loaded with the core action. Slightly packing mechanism + edge case into one sentence, but no wasted words.

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?

Given output schema exists (so return structure needn't be described), the description covers purpose, internal dependency, extraction logic, and key limitation (engine coverage). Complete for an agent to invoke correctly.

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 coverage is 100% and each parameter has a clear description. The description reinforces the engine limitation ('engines without raw_answer') which aligns with the schema's enum description, adding no new semantic detail beyond what the schema provides. Baseline 3 is appropriate.

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?

Specific verb+resource (extract cited snippet from raw answer for each citation). Explicitly distinguishes from the sibling citations_check by stating it 'Calls check_citations, then...' and clarifies the why-vs-that distinction.

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?

Gives a clear usage framing ('Use to see why an engine cited a URL, not just that it did') and notes an important limitation (returns 'not found' for Bing/Brave). However it doesn't explicitly contrast against the other citations_* siblings like citations_provenance.

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