skills/xobotyi/cc-foundry/ansible/Gen Agent Trust Hub

ansible

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external inventory and variable data that could potentially contain malicious input or instructions.
  • Ingestion points: YAML and INI inventory files, group_vars, host_vars, and external secret managers like HashiCorp Vault or AWS Secrets Manager as described in SKILL.md and inventory-management.md.
  • Boundary markers: No specific boundary markers are defined for the agent to distinguish between instructions and data within these sources.
  • Capability inventory: The skill utilizes powerful Ansible modules for arbitrary command execution (command, shell), file system operations (copy, template), and network communication (uri) across multiple referenced scripts.
  • Sanitization: The skill advocates for strong security practices such as using no_log: true to prevent credential exposure in logs and Ansible Vault for encrypting sensitive variables, detailed in vault-and-security.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 10:16 PM
Security Audit — agent-trust-hub — ansible