identity-trace
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill transmits identifiers (emails, phone numbers, usernames) to external OSINT services and tools. This behavior is documented and essential to the skill's primary purpose. It uses well-known, established services like Gravatar, HaveIBeenPwned, and GitHub for data collection.
- [COMMAND_EXECUTION]: The skill invokes external OSINT tools via the Python
subprocessmodule. It mitigates command injection risks by using list-based arguments and strictly normalizing all user-supplied inputs before they are passed to the shell. - [CREDENTIALS_UNSAFE]: The skill manages authentication tokens (HIBP_API_KEY, GH_TOKEN) via environment variables. The code includes specific logic to ensure these secrets are redacted from logs and diagnostic outputs.
- [PROMPT_INJECTION]: The skill processes untrusted data from OSINT sources and presents it to the agent. This represents an indirect prompt injection surface. The risk is minimized through sanitization (removal of control characters) and data normalization.
- Ingestion points: Tool outputs from holehe, maigret, and phoneinfoga, API responses from Gravatar, HIBP, and GitHub, and manual result ingestion via the ingest command.
- Boundary markers: None used in the generated markdown report.
- Capability inventory: Specific shell command execution of OSINT tools and HTTPS requests to defined providers.
- Sanitization: Uses clean_text to filter control characters and md_cell to escape markdown table delimiters.
Audit Metadata