trae-rules-writer
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill performs its stated purpose of generating IDE configuration rules based on project analysis without any malicious intent or hidden behaviors.
- [COMMAND_EXECUTION]: The skill executes a local Python script
scripts/init_rule.pyto initialize rules from templates. Analysis of the script shows it performs simple string replacement and file writing using standard library modules (argparse, pathlib). It does not access the network, use dynamic execution like eval(), or perform unsafe system operations. - [SAFE]: Project analysis performed by the
Project ScannerandConvention Detectoragents is restricted to reading project structure and code samples to extract style patterns. This data ingestion is necessary for the skill's primary function and occurs within the local project context. - [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it ingests untrusted data from the project codebase (READMEs, comments) to inform rule creation. Ingestion points: project files read by scanning agents. Boundary markers: absent. Capability inventory: local script file-write and agent read. Sanitization: absent. This is a low-risk vulnerability surface inherent to the skill's purpose of context-aware rule generation and does not indicate malicious intent.
Audit Metadata