feedback-analysis
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Risk.
- Ingestion points: The skill processes untrusted user-generated content via the
feedbackItems[].textfield as defined inschemas/input.schema.json. - Boundary markers: No clear delimiters or specific instructions exist within
SKILL.mdto prevent the agent from interpreting instructions that might be embedded in the feedback text. - Capability inventory: The skill metadata in
SKILL.mdexplicitly requests access to powerful tools includingBash,Write,WebSearch, andWebFetch. - Sanitization: While
hooks/pre-execute.cjsperforms structural validation of the input using theajvlibrary, it does not include content-level sanitization or filtering to mitigate malicious instructions within the feedback text.
Audit Metadata