semantic-coding
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Analysis identified a surface for indirect prompt injection, common in tools that process external source code. 1. Ingestion points: Reads .tsx, .jsx, .ts, and .js files from a user-specified path as described in SKILL.md. 2. Boundary markers: Content is read directly without specific delimiters or instructions to ignore embedded commands. 3. Capability inventory: The skill has file system read and write access for code refactoring. 4. Sanitization: Content is processed for style matches without prior sanitization. Conclusion: The risk is negligible as the agent follows deterministic refactoring rules rather than executing logic from the processed files.
- [SAFE]: The skill does not perform any network operations, download external code, or access sensitive system configuration files.
Audit Metadata