orchestration-best-practices
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill uses mandatory language and 'command' verbs to override default agent behavior. Specifically, it dictates that 'Claude MUST follow all 10 rules' and 'Claude MUST include checklist verification' in its responses.
- [PROMPT_INJECTION]: The 'Required Plugins' section mandates the use of a 'Superpowers plugin' and instructs the agent to 'prefer it over the default approach' for all tasks, which constitutes a behavioral steer away from standard safety or operational protocols.
- [PROMPT_INJECTION]: The 'Checklist: Before Finishing' section contains directives that force the agent to modify its output format and content based on the skill's specific requirements, overriding the agent's internal response generation logic.
Audit Metadata