The Agent Skills Directory

[SAFE]: The skill is designed to interact with the atris.ai API, which matches the skill author's verified identity. All network communications are directed to atris.ai and api.atris.ai domains.
[DATA_EXPOSURE]: The skill accesses an authentication token stored in ~/.atris/credentials.json. This is the documented credential location for the AtrisOS platform and is used solely to authenticate requests to the vendor's API.
[COMMAND_EXECUTION]: The skill utilizes shell commands (curl and node) to interact with the web API and parse configuration data. These commands are standard for CLI-based agent skills and are restricted to the skill's defined functionality.
[INDIRECT_PROMPT_INJECTION]: The skill reads and processes data from external Google Slides presentations, which may contain untrusted content.
Ingestion points: Presentation data is fetched via GET /presentations/{id} as described in SKILL.md.
Boundary markers: No explicit delimiters or warnings for the agent are present in the provided instructions.
Capability inventory: The skill has the ability to perform network requests via curl and can export content to PDF.
Sanitization: There is no evidence of content sanitization in the provided command examples; the skill relies on the agent's native handling of processed text.

slides