moai-project-language-initializer
Pass
Audited by Gen Agent Trust Hub on Mar 2, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted data from local configuration files and user inputs.
- Ingestion points: Processes data from
.claude/settings.json,.moai/config/config.json, and direct user input for nicknames and domain selections. - Boundary markers: No explicit delimiters are specified in the provided markdown for wrapping interpolated inputs, although the skill logic is designed to produce structured JSON configuration.
- Capability inventory: The skill is authorized to use
Read,Write,Edit,MultiEdit,Glob,TodoWrite, andAskUserQuestiontools. - Sanitization: The implementation includes validation logic such as a 20-character limit for nicknames and JSON schema validation for configuration files.
- [DATA_EXPOSURE]: The skill facilitates the configuration of Figma Access Tokens. It adheres to security best practices by guiding the user to store these credentials in environment variables or
.envfiles rather than hardcoding them within the project configuration. It also accesses the local.claude/settings.jsonfile to detect existing MCP server configurations for Figma, which is consistent with its primary purpose as an initializer. - [COMMAND_EXECUTION]: The skill utilizes the
AskUserQuestiontool to implement a batch-questioning pattern, which optimizes user interaction by reducing the number of turns required for setup.
Audit Metadata