guard
Pass
Audited by Gen Agent Trust Hub on Nov 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill configures PreToolUse hooks to execute external shell scripts (check-careful.sh and check-freeze.sh) located in sibling directories relative to the plugin root.
- [DYNAMIC_EXECUTION]: The skill uses eval to execute the output of a local binary script (~/.claude/skills/gstack/bin/gstack-paths), which is a common but dynamic pattern for environment configuration.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from the user that is interpolated into shell commands.
- Ingestion points: A directory path is requested via AskUserQuestion in SKILL.md.
- Boundary markers: None present for the user-provided input.
- Capability inventory: The skill utilizes the Bash tool to perform file operations and path resolution.
- Sanitization: While the instructions use double quotes for shell interpolation, the logic remains a potential surface for command injection if the agent does not properly escape the user's input.
Audit Metadata