tables-value-trace
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external table rows which can be a vector for indirect prompt injection. 1. Ingestion points: The skill retrieves row data including cell values, errors, and fields via the clay tables rows get command. 2. Boundary markers: The instructions do not define clear boundaries or provide instructions for the agent to ignore potentially malicious content within cell data. 3. Capability inventory: The skill utilizes Bash(clay *) and Bash(jq *) tools to interact with the environment. 4. Sanitization: No explicit sanitization, validation, or filtering of the ingested table data is performed before it is used in the dependency tracing logic or final report.
Audit Metadata