daytona-secrets-volume

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill manages sensitive API keys (e.g., OpenAI, Anthropic) and defines specific file paths for secret storage such as /daytona-secrets/*.env and .env files. It includes instructions to source these credentials into the active shell for testing purposes.
  • [COMMAND_EXECUTION]: The skill utilizes the daytona CLI to execute shell commands within sandbox environments. This involves process control (e.g., pkill) and the execution of setup scripts in the .devcontainer/ directory to facilitate environment configuration.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it sources data from external environment files (/daytona-secrets/*.env) into an environment where it has powerful shell execution capabilities via daytona exec. While it warns against printing values, it lacks formal boundary markers or input sanitization for the sourced data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 12:27 PM
Security Audit — agent-trust-hub — daytona-secrets-volume