agent-fleet-operations

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill facilitates the management of the agent fleet through the execution of system utilities and vendor CLI tools, including systemctl for service management, git for repository synchronization, uv for dependency management, and the pj and hermes commands for fleet operations.
  • [DATA_EXPOSURE]: Detailed workflows are provided for identifying and migrating credentials from potentially insecure locations like environment files and Git history into secure storage via DeLoSecrets and 1Password (op:// references). This is a defensive security practice.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with vendor-controlled infrastructure for operational tasks, such as verifying TTS voices at https://vox.delo.sh and updating agent code from the author's official GitHub repositories (github.com/delorenj). These are within the author context and are documented for intended functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes local configuration data from files such as .project.json and agents-registry.yaml. It mitigates potential injection or corruption risks through the use of strict schema validation, atomic file replacements, and kernel-level file locks during mutation windows.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:55 AM
Security Audit — agent-trust-hub — agent-fleet-operations