AGENT LAB: SKILLS

copilot-sdk

Warn

Audited by Snyk on Feb 15, 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). Yes — the SDK explicitly supports loading custom agent Markdown files (e.g., .github/agents/ in repositories or org-level agents) and exposes tools/MCP integrations including a "web" tool and MCP servers that can fetch repository or web content, so the agent can read untrusted, user-provided third-party content at runtime.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 15, 2026, 08:11 PM