sub-agent-patterns

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill provides documentation, architectural patterns, and prompt templates for utilizing Claude Code sub-agents.\n- [COMMAND_EXECUTION]: The documentation contains illustrative examples of standard development shell commands (e.g., git, npm, wrangler) used to explain sub-agent workflows and context management.\n- [DYNAMIC_EXECUTION]: The guide documents the use of lifecycle hooks to trigger local scripts (e.g., ./scripts/run-linter.sh) as part of an automated development workflow.\n- [PROMPT_INJECTION]: The skill provides guidance on writing effective system prompts for sub-agents, including using instructions like 'FIX issues found' to empower specialized agents. These are architectural recommendations for the platform and do not constitute malicious injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 08:44 AM
Security Audit — agent-trust-hub — sub-agent-patterns