Skip to main content
Glama
john-walkoe

USPTO Patent Citation MCP Server

by john-walkoe

Citations_search_oa_citations_balanced

Citations_search_oa_citations_balanced
Read-only

Search office action citations with full analytical detail, including statutory basis (102/103/112) and rejection type, to analyze prior art cited against patent applications.

Instructions

Search Office Action Citations (v2) with all 16 available fields. Prior art cited against an application, Form 892, Form 1449, IDS references, 102 103 112 statutory basis, action type, paragraph number.

Use after Citations_search_oa_citations_minimal for detailed analysis of selected applications. All fields: patentApplicationNumber, groupArtUnitNumber, techCenter, referenceIdentifier, parsedReferenceIdentifier, actionTypeCategory, legalSectionCode, examinerCitedReferenceIndicator, applicantCitedExaminerReferenceIndicator, officeActionCitationReferenceIndicator, workGroup, paragraphNumber, createDateTime, createUserIdentifier, obsoleteDocumentIdentifier, id.

This tier is where the OA-only analytical fields live: legalSectionCode (102/103/112 statutory basis) and actionTypeCategory ('rejected') have no equivalent in the enriched lane, and paragraphNumber locates the citation within the office action.

⚠️ APPLICANT-CITED (1449/IDS) COVERAGE IS PARTIAL. This lane is documented upstream as transcribing Form 892 AND Form 1449, but on IDS-heavy files it returns close to what the examiner applied and little else, in every era. Measured against the patents' own References Cited pages (union of BOTH lanes): US 7,971,071 -> 5 of 91 US 9,496,922 -> 1 of 251 US 9,135,462 -> 0 of about 620 US 11,656,067 -> 3 of 15, prosecuted 2021-2023 INSIDE the documented window A reference's absence here is NO evidence that the applicant did not disclose it, and a count from this lane is not the applicant's full IDS. For a complete 1449 record, read the IDS documents through the PFW MCP.

CROSS-LANE JOIN KEY: every row carries referenceKey, the normalised reference identifier, and it is the ONLY correct key for unioning this lane with the enriched lane. On app 12849948 this lane's parsedReferenceIdentifier reads '20060075466' while the enriched citedDocumentIdentifier reads 'US 2006/0075466 A1'; joining those two raw fields finds zero overlap on every application, when the true answer there is four references in both lanes. referenceKey is digits only (a leading US, spaces, slashes, hyphens and the kind code stripped, series markers such as RE kept) and is carried on both lanes at every tier. It is null on a row whose identifier does not reduce to a document number.

OA Citations v2 documented window: office actions mailed 2017-10-01 to ~30 days prior to today (older records have been observed in practice). No office-action date field exists — do not add an officeActionDate clause (HTTP 400).

⚠️ publicationNumber IN criteria IS A DELIBERATE 400 HERE, AND THAT IS A FEATURE. The raw upstream API answers that field with HTTP 200 and numFound 0, which reads as "this patent was never cited" and is silently wrong. This server refuses the clause so the mistake is visible. Use the patent_number parameter for the subject patent, or parsedReferenceIdentifier to find where a patent was CITED.

IDENTIFIERS: application_number is the APPLICATION serial. This index has no patent-number field (publicationNumber returns HTTP 400), so patent_number is crosswalked here: pass a GRANTED patent number (7-8 digits; commas, spaces and a US prefix accepted) and it is resolved to its application serial with one USPTO ODP applications-search call, then queried as patentApplicationNumber. The response reports the mapping in patent_number_resolution {input, interpreted_as, resolved_application_number, source}. An 11-digit pre-grant publication number is refused here (use Citations_search_citations_balanced for those), an unresolvable number is a 400 naming the accepted forms, and a patent_number that disagrees with a supplied application_number is a 400 rather than a query that can only return zero.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
rowsNo
startNo
fieldsNo
art_unitNo
criteriaNo
tech_centerNo
patent_numberNo
examiner_citedNo
application_numberNo

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault

No arguments

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.0.0

TDQS

A4.8/5.0
Behavior5/5

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

Annotations only declare readOnlyHint=true. The description adds extensive behavioral context: partial applicant-cited coverage with concrete patent metrics, the cross-lane join key (referenceKey), the documented time window, the deliberate HTTP 400 for publicationNumber, and the patent_number crosswalk behavior. No contradictions.

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?

The description is front-loaded with purpose and placement, then organizes critical caveats into distinct sections. It is verbose, but each section addresses a unique concern (coverage warning, join key, window, 400 behavior, identifier rules) that materially affects correct use. The length is justified by the tool's complexity.

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?

Covers all essential operational details: the 16 fields, the unique analytical fields, the partial coverage caveat with evidence, the join key, the time window, the deliberate 400 behavior, and the patent_number resolution. With an output schema present, an agent has everything needed to call it correctly and interpret results.

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 description coverage is 0%, and the description compensates well for the ambiguous parameters: patent_number (crosswalk to application serial), application_number (application serial), parsedReferenceIdentifier (join key usage), and criteria (publicationNumber refusal). It does not explicitly explain rows, start, fields, art_unit, tech_center, or examiner_cited, but these are self-evident from names and defaults. Overall strong compensation for the high-stakes parameters.

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?

Clearly states it searches Office Action Citations (v2) with all 16 fields, and positions itself as the detailed tier after the minimal search. It distinguishes from siblings by naming the unique analytical fields (legalSectionCode, actionTypeCategory, paragraphNumber) and by referencing the enriched lane.

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

Usage Guidelines5/5

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

Explicitly says to use after Citations_search_oa_citations_minimal for detailed analysis, and gives alternatives: use Citations_search_citations_balanced for pre-grant publications, and use PFW MCP for complete 1449 records. Also warns about partial applicant-cited coverage, guiding when not to rely on this lane.

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