hex-security-basics
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines a potential attack surface by processing external data from webhooks and Hex run parameters.
- Ingestion points: Processes
req.bodyin the webhook verification function andinput_paramswithin theHexRunRequestSchemadefined inSKILL.md. - Boundary markers: The skill explicitly instructs the agent to 'Never include tokens, SQL, output, or raw events' in its output, which serves as a protective boundary against leaking processed data.
- Capability inventory: The skill is restricted to standard file operations (
Read,Write,Grep) and does not include capabilities for arbitrary command execution or network operations. - Sanitization: Implements robust sanitization using
zodfor schema validation and aredactHexLogfunction to strip sensitive fields from processed records.
Audit Metadata