The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The fetch_docs.py script fetches documentation data from https://www.dash-mantine-components.com/assets/llms.txt. This is a standard utility function that retrieves official technical references to assist with code generation and component search.
[COMMAND_EXECUTION]: The skill performs controlled shell operations using the subprocess module:
fetch_docs.py dynamically installs the httpx package via pip if it is not present in the environment.
component_search.py executes the local fetch_docs.py script to perform online searches. These operations are hardcoded or use argument lists, preventing arbitrary command injection.
[PROMPT_INJECTION]: The skill ingests documentation from an external URL which is then searched and provided to the agent as context. This creates a surface for indirect prompt injection if the remote documentation were to be compromised, although the current source is the official technology domain.
[SAFE]: No malicious patterns such as credential theft, multi-layer obfuscation, or persistence mechanisms were detected. File system activities are limited to project scaffolding and standard caching within the user's local directory.

dmc-py