The Agent Skills Directory

[DATA_EXFILTRATION]: The skill documentation states that it "Automatically refreshes expired tokens using Google's cloud function." Official Google OAuth implementations refresh tokens directly via Google's authorized API endpoints (e.g., oauth2.googleapis.com). Using a third-party or developer-managed cloud function as an intermediary for token refreshing is a high-risk pattern that could allow the author to intercept or harvest sensitive OAuth 2.0 refresh and access tokens.
[COMMAND_EXECUTION]: The skill's primary functionality is implemented through external Python scripts (scripts/auth.py and scripts/slides.py). These scripts are executed via the command line with full access to the user's local environment. Since the scripts are not provided for analysis, their internal logic (including potential network exfiltration or filesystem access) cannot be verified.
[PROMPT_INJECTION]: The get-text command creates an indirect prompt injection vulnerability by ingesting arbitrary text from external Google Slides presentations into the agent's context.
Ingestion points: scripts/slides.py get-text (SKILL.md)
Boundary markers: None. The skill does not instruct the agent to ignore or delimit instructions found within the presentation text.
Capability inventory: The skill has full write access to Google Slides, including create, add-slide, replace-text, delete-slide, and batch-update capabilities.
Sanitization: There is no evidence of sanitization or filtering of the text extracted from slides before it is processed by the agent.

google-slides-automation