creativeclaw-submit-feedback
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied feedback, creating a surface where malicious instructions could be embedded in data.
- Ingestion points: User-provided bug reports and task descriptions processed in SKILL.md.
- Boundary markers: Lacks explicit delimiters for user input, though instructions direct the agent to synthesize concise reports.
- Capability inventory: Invokes the submit_feedback tool via the Creative Claw MCP interface defined in agents/openai.yaml.
- Sanitization: Instructions include a safety directive to exclude secrets, credentials, and personal information.
Audit Metadata