The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes shell scripts like generate-structure.sh and validate-skill.sh. Analysis shows these are local utility scripts used for scaffolding directory structures and performing static analysis (using grep, wc, and find) on the skill's documentation to ensure compliance with quality standards. No malicious or unauthorized command execution was detected.
[EXTERNAL_DOWNLOADS]: The skill's documentation (e.g., PYTHON-VERSION-MATRIX.md and WEB-RESEARCH.md) references various developer tools such as pytest, ruff, and uv, along with official documentation sites (e.g., python.org, pytest.org). These are standard technology resources and well-known services. The skill uses WebFetch and WebSearch to retrieve legitimate technical documentation to inform the skill generation process.
[DATA_EXFILTRATION]: While the skill requests the WebFetch tool, its usage patterns involve retrieving public documentation and best practices. There is no evidence of the skill attempting to access sensitive files (like SSH keys or environment variables) or sending private data to external servers.
[PROMPT_INJECTION]: The skill's instructions are focused on orchestrating the skill creation workflow. It does not contain patterns attempting to override the agent's core safety guidelines or extract system prompts. The 'Automatic Triggers' and 'Manual Invocation' sections are used for intent recognition related to the skill's primary purpose.
[INDIRECT_PROMPT_INJECTION]: The skill acts as an orchestrator that ingests data from external sources (WebSearch, WebFetch, Context7) to generate new content.
Ingestion points: Data enters via the WebFetch and WebSearch tools, as well as the mcp__context7__get-library-docs tool.
Boundary markers: The skill emphasizes the use of CHECKLIST.md and explicit validation steps to review generated content.
Capability inventory: Capabilities include Bash, Write, Edit, and Glob for file management.
Sanitization: The skill provides a validate-skill.sh script and a quality validation checklist to inspect generated artifacts for sensitive data or unsafe scripts before deployment. The risk is managed through these developer-focused validation frameworks.

moai-cc-skill-factory