The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script scripts/check_links.py to perform its analysis. The execution is triggered by user-initiated commands and targets files within the local project environment.
[EXTERNAL_DOWNLOADS]: The script uses the standard requests library to perform HTTP HEAD and GET requests. This is done solely to verify the reachability of external URLs found within Markdown documents. It includes a standard User-Agent header and handles common HTTP status codes.
[DATA_EXFILTRATION]: While the skill accesses the network, it does not transmit sensitive information. It only processes Markdown files to extract URLs and checks their availability, which is the primary and intended function of the tool.
[PROMPT_INJECTION]: The skill processes external data (Markdown files) which constitutes an indirect injection surface. However, the script includes logic to ignore links within code blocks, reducing the risk of accidental execution of example instructions. The functionality is limited to link validation and does not involve re-interpolating file content into LLM prompts in an unsafe manner.

md-link-checker