moonbit-agent-guide
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the
moonCLI toolchain for project management, compilation, and testing. This includes semantic code navigation commands such asmoon ide doc,outline, andrename, which are standard for language-specific agent skills.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines patterns for processing project source code and documentation, which constitutes an ingestion surface for untrusted data.\n - Ingestion points: Project source files (
.mbt), documentation tests (README.mbt.md), and API discovery outputs frommoon ide doc.\n - Boundary markers: Employs
///|as a separator for top-level code blocks and triple-backtick markdown blocks for embedded tests.\n - Capability inventory: The agent is authorized to perform file operations, execute build/test subprocesses, and manage dependencies.\n
- Sanitization: Relies on the MoonBit compiler's checked error system and snapshot-based test validation to ensure integrity.\n- [EXTERNAL_DOWNLOADS]: Installation instructions in
README.mdinvolve cloning the skill's source from the vendor's official GitHub repository (github.com/moonbitlang/moonbit-agent-guide).
Audit Metadata