choosing-swarm-patterns
Pass
Audited by Gen Agent Trust Hub on May 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a purely instructional document and does not contain any executable scripts, binary files, or network-enabled code.
- [PROMPT_INJECTION]: The skill provides templates for multi-agent workflows that interpolate external data into prompts (e.g.,
{{task}}). While this represents the standard attack surface for indirect prompt injection, there are no malicious instructions embedded in the skill itself. - Ingestion points: Variable interpolation in
SKILL.mdYAML workflow section ({{task}},{{steps.plan.output}}). - Boundary markers: Absent in provided examples.
- Capability inventory: Described agents utilize standard LLM interfaces (
cli: claude). - Sanitization: Not present in the provided templates.
Audit Metadata