preflight

Pass

Audited by Gen Agent Trust Hub on Sep 10, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where untrusted data from a user request is passed to specialized sub-agents, creating a surface for indirect prompt injection attacks.
  • Ingestion points: SKILL.md (Step 2: Independent Risk Challenge and Step 5: Verify with Specialists) requires passing the 'raw request' and 'evidence' directly into prompts for child agents.
  • Boundary markers: The instructions lack explicit data delimiters (such as XML tags or unique separators) to isolate the untrusted user payload from the instructions provided to the sub-agents.
  • Capability inventory: The parent agent maintains capabilities for code implementation (file system writes) and repository verification checks (shell command execution).
  • Sanitization: The workflow does not describe any sanitization, filtering, or validation of the raw request content before it is interpolated into sub-agent prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 10, 2026, 11:47 PM
Security Audit — agent-trust-hub — preflight