review
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it acts upon potentially malicious instructions embedded in pull request comments or CI logs. \n
- Ingestion points: Fetches data via 'gh pr view', 'gh run view', and repository scripts that list unresolved review threads. \n
- Boundary markers: The skill does not define specific boundaries or ignore-directives for the data being parsed. \n
- Capability inventory: The agent has access to 'Bash' and 'Edit' tools, allowing it to execute commands and modify the codebase based on the received input. \n
- Sanitization: No data sanitization or validation routines are specified for the processed content. \n- [COMMAND_EXECUTION]: The skill utilizes the 'Bash' tool to execute local scripts like '.claude/scripts/review-resolve.sh' and interact with CLI tools like 'gh' and 'forge'. While these are operational requirements for the skill's purpose, they represent the execution of logic that could be influenced by injected instructions.
Audit Metadata