synthesis-patterns

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
  • [Indirect Prompt Injection] (MEDIUM): The skill is designed to ingest and process content from external, untrusted sources which provides a surface for indirect prompt injection attacks.
  • Ingestion Points: The skill explicitly utilizes search, meeting_lookup, and code_search to pull data into the context (SKILL.md).
  • Boundary Markers: The instructions do not define delimiters or provide 'ignore embedded instructions' prompts to prevent the agent from executing commands hidden within the ingested documents or transcripts.
  • Capability Inventory: The skill enables access to highly sensitive organizational data including source code, employee information, and meeting records.
  • Sanitization: There is no evidence of sanitization or validation of the external content before it is interpolated into the synthesis output.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 10:39 AM