outside-to-issue
Pass
Audited by Gen Agent Trust Hub on Jun 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local shell script
scripts/transform-outside-to-issues.shto perform file operations and directory management. These scripts are assumed to be part of the local environment provided by the author. - [PROMPT_INJECTION]: The skill facilitates an indirect prompt injection surface by processing and reformatting untrusted review data.
- Ingestion points: Review content is read from Markdown files in the
ai-docs/reviews-pr-<PR>/outside/directory. - Boundary markers: No delimiters or instructions are added to the output to prevent downstream agents from obeying instructions embedded within the review body.
- Capability inventory: The resulting issue files are intended for use by tools such as
read_pr_issues.shand thepr-fixcommand, which may perform automated modifications or network actions based on the content. - Sanitization: There is no evidence of sanitization or escaping of the input content before it is written into the new issue files.
Audit Metadata