super-swarm-spark

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill parses untrusted markdown plan files and interpolates task descriptions into subagent prompts, creating a surface for indirect prompt injection. Ingestion points: plan file (Step 2); Boundary markers: absent; Capability inventory: file read/write, git commit, and test execution (Step 6); Sanitization: absent. This behavior is standard for development orchestration tools.
  • [COMMAND_EXECUTION]: The orchestrator includes steps to run tests and validation scripts during the integration phase. This is a standard capability for development automation tools to ensure code quality after parallel modifications and relies on the integrity of the project files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 12:43 PM