github-copilot-cli
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents how to process untrusted data from multiple sources, which presents a surface for injection attacks.
- Ingestion points: The skill utilizes the '@' syntax to read local files and integrates with Model Context Protocol (MCP) servers to fetch data from GitHub issues, PostgreSQL databases, and Brave search results.
- Boundary markers: No specific boundary markers or delimiters for untrusted data are specified in the provided instructions.
- Capability inventory: The skill enables the agent to suggest shell commands, generate and refactor code, and interact with external systems via MCP tools.
- Sanitization: No explicit sanitization or validation of external content is mentioned in the documentation.
- [EXTERNAL_DOWNLOADS]: The skill includes instructions to fetch the GitHub CLI and its Copilot extension from official GitHub domains. These references target well-known and official infrastructure.
- [COMMAND_EXECUTION]: The documentation provides commands for system installation on Linux, including the management of GPG keys and package repositories. These are standard administrative tasks for installing the official CLI tool.
- [REMOTE_CODE_EXECUTION]: The skill describes the use of 'npx' to dynamically load and execute official Model Context Protocol servers for extended functionality. These packages originate from established and recognized repositories.
Audit Metadata