The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is centered around the mcp_imagekit_api_execute tool, which requires the agent to generate and run TypeScript code dynamically to perform operations such as file management, folder manipulation, and cache invalidation. This pattern represents a significant execution surface where the agent's generated code acts as the primary driver for system interactions.
[COMMAND_EXECUTION]: The instructions explicitly mandate the use of an external script (skills/upload-files/resources/upload.py) for handling local file uploads, as the main API execution tool is limited to URL-based uploads. This involves the execution of code logic stored in an external resource file.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through its use of the mcp_imagekit_devtools_search_docs tool.
Ingestion points: Data is ingested from the output of the RAG-powered search tool, which retrieves information from public documentation, guides, and community posts (SKILL.md).
Boundary markers: The skill lacks explicit boundary markers or instructions to treat tool outputs as data rather than instructions.
Capability inventory: The agent possesses powerful capabilities through the execute tool and the upload.py script (SKILL.md).
Sanitization: There are no documented sanitization, filtering, or validation steps for the content retrieved from external community and documentation sources.

mcp-preflight