get-to-know-a-client

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect prompt injection surface identified. The skill is designed to ingest and analyze untrusted external data from client notes and transcripts, which it then uses to generate technical architecture and executable code.
  • Ingestion points: The process in SKILL.md involves reading all available client information, notes, and meeting transcripts in Phase 1.
  • Boundary markers: The skill lacks explicit instructions or delimiters for the agent to ignore or isolate potentially malicious instructions embedded within the processed client documents.
  • Capability inventory: In Phase 4 of SKILL.md, the agent is instructed to provide exact commands, code snippets, and configuration files for the implementation plan.
  • Sanitization: No sanitization, validation, or filtering procedures are defined for the incoming client data before it is used to generate the build plans.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 02:53 PM
Security Audit — agent-trust-hub — get-to-know-a-client