reflect-reviews
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests review information from external messages to decide on code modifications and repository actions. This creates an attack surface where malicious instructions embedded in review comments could influence the agent's output.\n
- Ingestion points: External review information provided in user messages.\n
- Boundary markers: The skill lack specific delimiters or instructions to ignore potential commands embedded within the review data.\n
- Capability inventory: The skill is authorized to modify files, commit changes, and create pull requests.\n
- Sanitization: No sanitization or validation of the review content is specified before processing.\n- [COMMAND_EXECUTION]: The skill is instructed to perform git operations, including adding commits to existing branches and creating new regular or draft pull requests based on the review outcome.
Audit Metadata