The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill metadata specifies a dependency on an external MCP server @clawfu/mcp-skills. Loading external servers or extensions from unverified sources introduces risk to the agent environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input, such as business ideas and customer interview notes, to generate analysis. This creates a surface for indirect prompt injection attacks where malicious instructions could be embedded in the provided data.
Ingestion points: The Analyze Interview Responses and Validate a Business Idea sections in SKILL.md accept arbitrary user-pasted content for processing.
Boundary markers: The instructions do not define delimiters or provide explicit instructions for the agent to ignore commands embedded within the data provided for analysis.
Capability inventory: The skill is designed to perform text generation, formatting, and qualitative analysis within the agent's context.
Sanitization: No sanitization or validation of the input content is performed before the data is interpolated into the agent's prompt context.
[METADATA_POISONING]: The skill contains inconsistent author information, identifying as ClawFu in the frontmatter and MKTG Skills in the footer metadata, both of which conflict with the provided vendor context of guia-matthieu.

mom-test