The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the find utility for file discovery. It interpolates user-provided folder paths directly into shell commands (e.g., find "<folder>" ...), which creates a potential command injection vector if the user input is not properly sanitized or escaped before execution by the underlying shell tool.
[PROMPT_INJECTION]: There is an indirect prompt injection surface as the skill processes content from external Office documents and PDFs. Ingestion points: Document content is read during the audit process. Boundary markers: The skill does not define clear delimiters or instructions to ignore embedded prompts within the documents. Capability inventory: The agent can execute shell commands (find), write files to the local directory, and generate automation scripts. Sanitization: No sanitization of document-extracted text is specified, meaning malicious instructions in a document could influence the orchestration logic.
[EXTERNAL_DOWNLOADS]: The skill provides CI/CD integration templates that reference external resources, specifically official GitHub Actions such as actions/checkout@v4 and actions/upload-artifact@v4. These are well-known services and the references are documented neutrally for user configuration purposes.

document-accessibility-wizard