guided-flow-review
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from GitHub pull request titles, bodies, and code changes which could potentially contain malicious instructions meant to subvert the agent's behavior. Ingestion points occur when fetching data via
gh pr viewandgit diff. The skill mitigates this by instructing the agent to treat new or modified instruction files as code under review rather than active instructions. It maintains a manual-only execution mode and requires explicit user confirmation gates before recording findings or performing fixes. - [COMMAND_EXECUTION]: The skill relies on various system and project-local commands to perform its duties. It executes Git and GitHub CLI commands to identify changes and project-specific commands for formatting, linting, and testing. While these are standard for a development tool, they represent an execution of local code triggered by the skill's workflow.
Audit Metadata