brand-name-checker
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data from web search results and public API responses (e.g., from npm, PyPI, and WIPO). This data ingestion is inherent to the skill's function. The architecture mitigates risk by requiring subagents to return structured JSON and instructing the synthesizer agent to treat worker evidence as untrusted, verifying all facts against a centralized risk policy before generating recommendations.
- [EXTERNAL_DOWNLOADS]: The skill interacts with reputable and official external sources to retrieve availability and trademark data. These include GitHub, Instagram, Twitter/X, npmjs.org, pypi.org, brew.sh, wipo.int, and euipo.europa.eu. All network operations are conducted using standard platform tools (WebSearch, WebFetch) and target well-known, safe domains necessary for the brand checking task.
Audit Metadata