merge
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs shell-based operations including Git commands (
checkout,merge,tag, andbranch -D) to manage project history and execution of a local Python scriptsession_manager.pyfor worktree cleanup and state management. These actions are directly aligned with the skill's stated purpose.- [INDIRECT_PROMPT_INJECTION]: The skill processes session data such as task descriptions and agent identifiers which are interpolated into Git commit messages and summary files. While this creates a theoretical surface for injection if session metadata is untrusted, the impact is minimized by the skill's specialized context. - Ingestion points: Variable placeholders (
{session-id},{winner},{task}) inSKILL.mdare interpolated into Git commands and generated markdown files. - Boundary markers: No explicit delimiters or boundary instructions are provided in the command templates.
- Capability inventory: The skill has capabilities for file system modification (via Git and summary file writing) and local script execution.
- Sanitization: The instruction template does not demonstrate explicit escaping or validation of session-specific variables before interpolation.
Audit Metadata