plan-executor
Warn
Audited by Socket on Feb 26, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
The code fragment depicts a coherent orchestration tool for plan-based multi-agent execution, with clear separation of duties, user confirmation gates, and per-task SubAgent prompts. It exhibits no obvious malicious activity, credential handling, or external data exfiltration within the fragment. The primary concerns are prompt/data exposure through SubAgent prompts and potential template staleness or plan non-conformity, which are manageable through existing safeguards (confirmation step, template design, and validation). Overall, a sound, purpose-aligned orchestration component with moderate operational risk centered on prompt/data exposure and dependency parsing robustness.
Confidence: 75%Severity: 75%
Audit Metadata