The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides Python code snippets for interacting with the Fathom and Maton APIs. These scripts use the standard urllib.request library and are intended to allow the agent to perform meeting-related tasks on behalf of the user.
[DATA_EXFILTRATION]: The skill performs network operations to maton.ai domains (gateway.maton.ai, ctrl.maton.ai, connect.maton.ai). This communication is required for the skill's primary purpose of OAuth management and Fathom API access. No unauthorized data transmission or sensitive local file access was found.
[CREDENTIALS_UNSAFE]: The skill follows best practices by requiring the use of an environment variable (MATON_API_KEY) for authentication rather than hardcoding secrets in the skill body.
[PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface by processing external meeting transcripts and summaries. While this is an inherent characteristic of meeting-analysis tools, the skill itself does not contain malicious instructions or bypass techniques.

fathom