llm-wiki-news-radar
Pass
Audited by Gen Agent Trust Hub on Aug 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from various external web sources, such as GitHub repositories, research papers, and community blogs. Because it also provides a mechanism to write this content to the local file system (via the 'file-back' procedure to a wiki vault), there is an inherent risk that malicious instructions hidden within those external sources could be executed by the agent or persisted to the user's environment.
- Ingestion points: External web content from various domains identified during search.
- Boundary markers: The skill lacks explicit instructions or delimiters to warn the agent to ignore instructions embedded within the retrieved external content.
- Capability inventory: The skill utilizes web browsing and has the capability to write files to the 'wiki/queries' directory.
- Sanitization: There is no evidence of content sanitization or validation of external data before it is formatted and saved to the local file system.
Audit Metadata