workspace-collab

Fail

Audited by Gen Agent Trust Hub on Apr 22, 2026

Risk Level: HIGHCOMMAND_EXECUTIONDATA_EXFILTRATIONREMOTE_CODE_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a shell command template that uses ssh and docker exec to perform administrative tasks on a remote server. It also encourages using the SECURITY DEFINER attribute in SQL functions, a pattern that grants elevated privileges to those functions regardless of the caller's permissions.\n- [DATA_EXFILTRATION]: The skill contains a direct instruction to run a command referencing a sensitive private key at ~/.ssh/id_ed25519. Accessing such credential files within the context of network-enabled commands like ssh creates a risk of exfiltration or unauthorized access.\n- [REMOTE_CODE_EXECUTION]: By instructing the agent to run commands on a remote VPS (defined by variables like ${VPS_HOST}), the skill establishes a remote code execution vector that could be exploited if the target host or connection parameters are manipulated.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Apr 22, 2026, 01:14 AM
Security Audit — agent-trust-hub — workspace-collab