content-repurposer
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process long-form content from untrusted external sources, including URLs and local files, without specific boundary markers or instructions to ignore embedded commands.
- Ingestion points: The skill accepts content via pasted text, file paths, or URLs in 'Step 1: Gather Inputs' of
SKILL.md. - Boundary markers: There are no delimiters or instructions to treat the ingested content as data rather than instructions (e.g., no 'ignore instructions inside this content' warnings).
- Capability inventory: The agent has the capability to read local files, fetch remote content via URL, and generate structured output based on that content.
- Sanitization: No sanitization or validation of the input content is specified before it is processed by the agent's logic.
- [DATA_EXFILTRATION]: The skill explicitly allows for the ingestion of data via 'file path'. While intended for content like newsletters or research docs, this feature could be used to target sensitive local files (e.g., configuration files or credentials) if a user provides an unintended path, which the agent will then summarize and present in its output.
Audit Metadata