uspto-database

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the installation of standard Python packages (requests, pandas, matplotlib, google-cloud-bigquery) and connects to well-known data services including api.patentsview.org and Google Cloud BigQuery. These operations are consistent with the skill's stated purpose of retrieving patent metadata.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted text data (patent titles, abstracts, and claims) from the USPTO and Google Patents databases. This introduces a vulnerability surface where malicious instructions could theoretically be embedded in patent records to influence the agent's behavior during processing.
  • Ingestion points: Data enters the agent's context through requests.post calls to the PatentsView API and BigQuery queries in SKILL.md.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the processing logic.
  • Capability inventory: The skill uses requests for network access and matplotlib/pathlib for file system writes (saving charts and caching JSON).
  • Sanitization: The provided code snippets perform data frame manipulation and plotting but do not include explicit sanitization or filtering of the text content fetched from the APIs.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 12:54 AM