seed-deprecation
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides legitimate instructions for a design system maintenance workflow, including code documentation updates and synchronization between source files and user-facing guidelines. No evidence of malicious behavior, data exfiltration, or unauthorized privilege escalation was found.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied information such as deprecation reasons and replacement suggestions to be interpolated into source code comments and documentation. This represents a standard surface for indirect injection where malicious text could be embedded into the codebase, though the risk is low and inherent to the task's purpose.
- Ingestion points: User-provided inputs for deprecation details (Target, Replacement, Reason) and existing project documentation (deprecations.mdx).
- Boundary markers: None mentioned in the instructions for separating user input from the generated JSDoc/YAML content.
- Capability inventory: The skill allows the agent to modify source code files (.tsx, .yaml) and documentation (.md, .mdx), and execute internal code generation commands mentioned in the project's root documentation.
- Sanitization: No explicit sanitization or validation of the input strings is defined before they are written to files.
Audit Metadata