panews
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEOBFUSCATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [OBFUSCATION]: The primary execution script,
scripts/cli.mjs, is provided in a minified format. This practice obscures the underlying logic and prevents a clear security audit of the tool's behavior, although it appears to be a bundled version of the tool and its validation libraries. - [INDIRECT_PROMPT_INJECTION]: The skill fetches news articles, editorial rankings, and community comments from external PANews and Polymarket APIs, which are then processed by the agent. This represents an indirect prompt injection surface where instructions embedded in news content or user comments could attempt to influence the agent.
- Ingestion points:
scripts/cli.mjsretrieves data fromuniversal-api.panewslab.comandpolymarket-boards.panewslab.com. - Boundary markers: The skill does not employ explicit delimiters or boundary markers to separate untrusted external content from the agent's instructions in the workflow references.
- Capability inventory: The skill allows the agent to perform network requests, search for articles, and read full article contents through the
scripts/cli.mjstool. - Sanitization: There is no evidence of sanitization, filtering, or escaping of the fetched API content before it is presented to the agent in Markdown format.
Audit Metadata