projectsetup
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes standard project setup commands such as
npm installfor testing frameworks andclaude mcp addfor configuring tool servers. It also references local utility scripts located in the user's home directory (~/.claude/scripts/) which are part of the intended developer toolset environment. - [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the well-known
@playwright/testpackage and associated browser binaries. These downloads originate from trusted registries and official project repositories. - [PROMPT_INJECTION]: The skill creates a project-level configuration file (
CHECKPOINTS.md) that explicitly defines the agent's autonomy levels. This includes 'AUTO-PROCEED' permissions for file editing and directory creation. While this modifies default agent behavior, it is implemented as a transparent project principle for efficiency rather than a malicious bypass of safety guidelines.
Audit Metadata