The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/attach_comment_files.py executes system commands including gh (GitHub CLI) and npx @playwright/cli. These commands are invoked using subprocess.run with a list of arguments, which is a secure method that prevents shell injection vulnerabilities.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx to run the Playwright CLI, which fetches the package from the official NPM registry. It also interacts with GitHub and GitHub Enterprise Server domains to facilitate the file upload process. These are recognized as well-known and trusted services.
[INDIRECT_PROMPT_INJECTION]: The skill provides a surface for indirect prompt injection because it allows an agent to process local file paths for upload to the cloud.
Ingestion points: The files argument in scripts/attach_comment_files.py accepts multiple local file paths from the agent's input context.
Boundary markers: None. The script does not include explicit instructions or markers to distinguish between trusted and untrusted file paths provided by an agent.
Capability inventory: The script performs file system operations (reading and copying files in stage_files) and utilizes browser automation with network access to upload these files to GitHub.
Sanitization: The script resolves and expands file paths using standard Python pathlib functions but does not implement a sandbox or restricted directory list for these operations.

shunk031-github-comment-attach-files