copilot-cli-quickstart
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts on untrusted user input to manage tutorial tracks and progress, creating a potential surface for indirect injection.
- Ingestion points: User responses to
ask_userprompts and natural language queries to determine intent (e.g., 'start tutorial', 'switch track'). - Boundary markers: No explicit delimiters or boundary instructions are used to separate user input from system logic.
- Capability inventory: The skill utilizes the
sqltool for state management andfetch_copilot_cli_documentationfor retrieving external data. - Sanitization: No explicit validation or sanitization rules for user input are defined in the instructions.
- [SAFE]: The skill references official GitHub documentation and tools (cli.github.com, docs.github.com) which are recognized as well-known, trusted services. The use of SQL for progress tracking is localized and does not involve exfiltration of sensitive data.
Audit Metadata