The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (LOW): The skill documentation recommends installing the ast-grep tool via Homebrew, npm (@ast-grep/cli), and Cargo. While these are reputable package managers, the ast-grep organization and repository are not included in the predefined trusted list. The severity is downgraded to LOW as these downloads are essential for the primary purpose of the skill.
[COMMAND_EXECUTION] (LOW): The skill instructs the agent to execute shell commands such as sg run and sg scan. These commands are used to process and transform local files, which is consistent with the skill's functionality.
[PROMPT_INJECTION] (LOW): This finding identifies a surface for Indirect Prompt Injection (Category 8).
Ingestion points: The ast-grep tool processes local source code files (e.g., src/, rules.yml) which may originate from untrusted external sources.
Boundary markers: Absent. The skill provides no instructions for the agent to use delimiters or ignore instructions within the analyzed code.
Capability inventory: The skill uses Bash to invoke sg for pattern matching and automated rewrites.
Sanitization: Absent. External code is processed and matched against AST patterns without pre-validation of the code's content or safety.
[CREDENTIALS_UNSAFE] (SAFE): Examples in modules/security-rules.md contain patterns like apiKey: "$$$KEY". These are structural templates for detection rules rather than actual hardcoded credentials.

moai-tool-ast-grep