playbook-executor

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs shell commands using a 'Run tool' to execute git add, git commit, and git push. This is used to synchronize generated Ansible playbooks with the user's remote repository as part of the mandatory 'Git Flow' workflow.
  • [REMOTE_CODE_EXECUTION]: The core functionality involves launching Ansible playbooks on remote systems via AAP job templates using the job_templates_launch_retrieve tool. This capability is mitigated by mandatory human-in-the-loop checkpoints before any execution begins.
  • [PROMPT_INJECTION]: The skill processes untrusted data from Ansible job logs and console output (jobs_stdout_retrieve) during Phase 5 (Execution Report). This represents an indirect prompt injection surface as the agent parses this data for CVE validation and status reporting.
  • Ingestion points: Job stdout and host summaries are retrieved from the AAP API and processed in Phase 5 of SKILL.md.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded commands are specified when the agent analyzes the stdout for CVE validation.
  • Capability inventory: The agent has the ability to execute shell commands (git) and launch additional jobs in the AAP environment.
  • Sanitization: There is no evidence of sanitization or filtering of the job logs before they are processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 06:05 PM