executing-plans
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill instructs the agent to read external plan files and 'Follow each step exactly'. If a plan file contains malicious instructions, the agent may execute them as part of its batch execution process.
- Ingestion points: SKILL.md (Step 1: 'Read plan file').
- Boundary markers: Absent; the skill does not define specific delimiters for plan content or instructions to ignore embedded prompts.
- Capability inventory: Implied shell command execution and file modification (Step 2: 'Execute Batch', 'Run verifications') and execution of choices (Step 5).
- Sanitization: Absent; the skill relies on the agent's 'critical review' and a human partner to detect issues before execution.
- [Prompt Injection] (SAFE): No direct attempts to override system prompts or bypass safety filters were detected within the skill text itself.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file paths, or network exfiltration patterns were identified.
- [Remote Code Execution] (SAFE): No external script downloads or package installations are performed by this skill.
Audit Metadata