skills/fastfold-ai/skills/boltz/Gen Agent Trust Hub

boltz

Pass

Audited by Gen Agent Trust Hub on Jul 30, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the official CLI tool using a shell script from the provider's domain (https://install.boltz.bio/boltz-api/install.sh). This is standard installation behavior for the service.
  • [REMOTE_CODE_EXECUTION]: The installation process involves piping a remote shell script to sh. This is documented in the SKILL.md under the Install section and targets the official service endpoint.
  • [COMMAND_EXECUTION]: The skill makes extensive use of the boltz-api command-line interface to perform scientific workflows. It also includes a persist.sh script to mirror artifacts between ephemeral and persistent storage in hosted environments.
  • [CREDENTIALS_UNSAFE]: The skill requires a BOLTZ_API_KEY to be set in the environment. It correctly advises users to use hosted environment injection or a local .env file, following secret management best practices rather than hardcoding credentials.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 30, 2026, 11:01 PM
Security Audit — agent-trust-hub — boltz