The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The authentication module scripts/skywork_auth.py and the API client scripts/excel_api_client.py access and manage a persistent token stored at ~/.skywork_token. Accessing sensitive credential files in the user's home directory is a high-risk pattern for data exposure.
[COMMAND_EXECUTION]: The authentication script scripts/skywork_auth.py uses subprocess.Popen to execute shell commands (open, start, xdg-open) for the purpose of launching a web browser during the OAuth login flow.
[EXTERNAL_DOWNLOADS]: The skill performs network operations and downloads files from api-tools.skywork.ai and api.skywork.ai. These domains are not recognized as trusted or well-known services and serve as the backend for the skill's data processing.
[PROMPT_INJECTION]: The skill processes untrusted data from external sources (Excel, CSV, PDF, and image files) which presents an indirect prompt injection risk.
Ingestion points: Files provided by the user and uploaded to the backend via client.upload_file in scripts/excel_api_client.py.
Boundary markers: The instructions do not implement specific boundary markers or delimiters for data being sent to the backend agent.
Capability inventory: The skill has the capability to write files to the local system, perform network requests to external APIs, and execute subprocesses.
Sanitization: There is no evidence of sanitization or validation of the content of the processed files within the client-side scripts.

skywork-excel