setup-toolchain
Pass
Audited by Gen Agent Trust Hub on May 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements security-focused hooks that monitor and restrict the agent's command execution within a specific project context.
- [COMMAND_EXECUTION]: The
enforce-toolchain.shscript acts as a defensive guard. It intercepts bash commands and uses regular expressions to block unauthorized tools (npm, npx, tsc) and destructive operations such as unrestricted recursive removal, hard git resets, and forced pushes. It uses standard platform protocols (exit codes and JSON output to stderr) to deny tool execution. - [SAFE]: The
session-env.shscript configures the agent's environment with project-specific variables and performs safe, automated cleanup of temporary session directories in/tmpusing patterns specific to the tool's operations. - [SAFE]: The instructions correctly guide the user to place scripts in project-local directories and update settings files, adhering to standard practices for AI agent extensions.
Audit Metadata