skills/involvex/skills/copilot-sdk/Gen Agent Trust Hub

copilot-sdk

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill documentation provides instructions to download and install packages from official registries and repositories associated with the GitHub organization.\n- [INDIRECT_PROMPT_INJECTION]: The implementation patterns for interactive assistants in SKILL.md create a surface for indirect prompt injection by processing raw user input directly through the SDK.\n
  • Ingestion points: Raw user input is collected using standard library functions like input(), readline, and os.stdin in the examples provided within SKILL.md.\n
  • Boundary markers: The sample code in SKILL.md lacks boundary delimiters or specific instructions to ignore instructions that might be embedded within the user input data.\n
  • Capability inventory: The skill facilitates capabilities for custom tool execution, session persistence, and connectivity to remote MCP servers as defined in SKILL.md.\n
  • Sanitization: The demonstration code does not include sanitization or validation logic for the ingested prompt content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 05:21 PM