Copilot Objection

Warn

Audited by Snyk on Mar 3, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (high risk: 0.80). The skill's mandatory Layer 2 "Enrichment (Web Search)" instructs the agent to run executeWebSearch queries (e.g., competitor claims, industry benchmarks, proof points) against open web sources and then use those results to shape the ACE response, which clearly ingests untrusted public content that can influence decisions and actions.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 3, 2026, 11:37 AM