The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The presentations read command extracts text and table data from external Google Slides presentations. This ingested content is untrusted and could contain malicious instructions designed to influence the agent's behavior. \n
Ingestion point: scripts/google-slides.py via the read_presentation_content and _extract_slide_text functions. \n
Capability inventory: The skill has extensive write capabilities, including creating and deleting slides and inserting text, shapes, and images, which could be misused if the agent follows instructions found in a presentation. \n
Sanitization: No explicit boundary markers or sanitization logic is used for the extracted text to separate instructions from data. \n- [EXTERNAL_DOWNLOADS]: The skill requires several standard Python libraries for Google API interaction and authentication. It also fetches images from user-supplied URLs to insert into slides as part of its primary functionality. \n
Dependencies: google-auth, google-auth-oauthlib, google-api-python-client, keyring, and pyyaml. \n
Image fetching: The images create command uses a user-supplied image-url to download and embed images. \n- [COMMAND_EXECUTION]: The skill executes a local Python script google-slides.py to handle all interactions with the Google Slides API. This is the intended purpose of the skill and is restricted to the specific script provided. \n- [SAFE]: Authentication is handled via OAuth 2.0. The skill follows security best practices by using the system keyring service to securely store and retrieve access tokens, rather than hardcoding credentials or using insecure local storage for tokens.

google-slides