browser-pilot
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill navigates to and extracts data from external websites. This untrusted content may contain instructions to redirect the agent's actions.
- Ingestion points: Web content retrieved through snapshots and console messages in SKILL.md.
- Boundary markers: No specific delimiters are mentioned to isolate untrusted data.
- Capability inventory: The skill can write files to the /output/ directory, navigate the network, and upload local files via the upload_file tool as described in SKILL.md.
- Sanitization: No sanitization of the extracted data is described.
- [DATA_EXFILTRATION]: The skill's documented capabilities include uploading local files to web forms and saving session cookies or authentication tokens to the local workspace. While intended for automation, these features represent potential vectors for data exfiltration if the agent follows malicious instructions from an untrusted webpage.
Audit Metadata