ds-swarm
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to partition tasks among multiple 'workers' and merge their outputs into a single report. This creates an attack surface where malicious instructions returned by a worker could influence the parent agent during the merging phase.\n
- Ingestion points: Results and evidence collected from worker sub-tasks (SKILL.md, Steps 7 and 8).\n
- Boundary markers: The instructions do not define specific delimiters or instructions to treat worker output as untrusted content to prevent accidental execution of embedded instructions.\n
- Capability inventory: The skill utilizes the ability to spawn subagents or execute briefs in fresh contexts (SKILL.md, Step 5).\n
- Sanitization: The instructions lack guidance for sanitizing or escaping worker-provided content before it is interpolated into the final merged report.
Audit Metadata