swarm-patterns

Pass

Audited by Gen Agent Trust Hub on May 25, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill describes orchestration patterns (e.g., Pattern 4) that interpolate untrusted content directly into sub-agent prompts, creating a vulnerability surface for Indirect Prompt Injection.
  • Ingestion points: Research results and file content are interpolated into agent prompts (e.g., ${research.content} in Pattern 4, and file lists in Pattern 3).
  • Boundary markers: The provided code snippets do not utilize XML tags, delimiters, or explicit instructions to the sub-agents to ignore potentially malicious content within the interpolated data.
  • Capability inventory: The agents used in these patterns include general-purpose and specialized reviewers which possess capabilities to write tasks, send messages, and potentially execute code or modify the file system depending on their environment.
  • Sanitization: There is no evidence of data sanitization or validation before interpolation into sub-agent prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
May 25, 2026, 01:54 PM
Security Audit — agent-trust-hub — swarm-patterns