dispatching-parallel-agents
Pass
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of instructional content and templates for effective agent orchestration. It does not contain executable code, remote downloads, or malicious patterns.
- [SAFE]: The skill emphasizes security-positive patterns, specifically context isolation (preventing agents from inheriting irrelevant conversation history) and filesystem isolation (using git worktrees to prevent overlapping file writes).
- [SAFE]: External references are limited to other skills within the author's own repository suite (krzysztofsurdy/code-virtuoso), which is consistent with the skill's purpose as a part of a larger agent toolkit.
- [SAFE]: The skill includes detailed guidance on preventing 'Indirect Prompt Injection' by mandating strict boundaries and structured output formats in subagent briefs, effectively reducing the risk of subagents being manipulated by untrusted data.
Audit Metadata