agnxi-search
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill retrieves sitemap data from the author's official domain (agnxi.com) to provide agents with up-to-date discovery data.
- [COMMAND_EXECUTION]: The tool runs the included Python script 'search.py' via a subprocess call to perform text-based filtering on the retrieved XML content.
- [SAFE]: No malicious patterns such as obfuscation, data exfiltration, or persistence were detected. The script utilizes only standard library components.
- [SAFE]: Indirect Prompt Injection Analysis: (1) Ingestion point: 'search.py' fetches 'https://agnxi.com/sitemap.xml'. (2) Boundary markers: The output uses standard bulleted list formatting without specific 'ignore instructions' delimiters. (3) Capability inventory: Includes local file execution of 'search.py' for parsing. (4) Sanitization: The script specifically targets and extracts URL locators from within XML tags, reducing the risk of processing unintended instructions.
Audit Metadata