idea-discovery
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data operations were detected. The skill primarily serves as a guide for product discovery processes.- [PROMPT_INJECTION]: Analysis of the instructions shows no attempts to override agent behavior, bypass safety protocols, or extract system prompts. The skill presents a surface for indirect prompt injection through the
/analyze-feedbackcommand, which processes user-provided data. Ingestion points: User feedback provided via the<file or paste>argument inSKILL.md. Boundary markers: None explicitly defined in the instructions. Capability inventory: The workflow involves writing results to the local file system (docs/ideas/,docs/insights/) and passing data to downstream evaluation tools. Sanitization: No validation of input content is specified. As this is a workflow definition, the surface is considered a feature of the intended data processing.- [DATA_EXFILTRATION]: No network operations or sensitive file access patterns were found. The skill describes standard local file operations for storing research results consistent with its stated purpose.- [REMOTE_CODE_EXECUTION]: There are no external dependencies, remote script downloads, or dynamic execution patterns present in the skill.- [COMMAND_EXECUTION]: The example commands provided illustrate intended workflow usage and do not involve high-risk shell operations or privilege escalation.
Audit Metadata