hypergrok-bootstrap
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches the HyperGrok trading desk components from the developer's official GitHub repository to initialize the workspace.
- Evidence: Clones 'https://github.com/galleonlabs/hypergrok-trading-desk.git' during the installation step.
- [COMMAND_EXECUTION]: Executes shell and Python scripts provided within the downloaded repository to perform system checks and retrieve public market data.
- Evidence: Runs 'bash scripts/check.sh', 'python3 scripts/opening_bell.py', and 'python3 scripts/desk_doctor.py'.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external agent profiles and shared skill instructions from the cloned repository to configure the trading desk interface.
- Ingestion points: Reads configuration data from the 'agents/' and 'skills/' directories within the 'hypergrok' workspace.
- Boundary markers: The skill includes a 'Safety boundary' section that explicitly restricts the agent from requesting wallet keys or initiating orders.
- Capability inventory: The skill utilizes subprocess execution for diagnostic scripts and file-write operations to organize the '/workspace/trading-desk' directory.
- Sanitization: Uses Git commit verification and specific repository structural checks before proceeding with the setup.
- [SAFE]: The skill instructions emphasize security best practices by enforcing a read-only engagement level and forbidding the provisioning of trading credentials.
Audit Metadata