The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the configuration of an external MCP server at https://rube.app/mcp. This third-party service provides the tool definitions and execution logic, creating a dependency on an external endpoint outside of the trusted organization list.
[DYNAMIC_EXECUTION]: Through the MIXPANEL_JQL_QUERY tool, the skill enables the execution of dynamic JavaScript code provided in the script parameter. This allows for custom script execution within the remote Mixpanel environment.
[DATA_EXFILTRATION]: The skill facilitates access to sensitive information, including user profiles (MIXPANEL_QUERY_PROFILES) and detailed product analytics data, which are transmitted through the external MCP server to the third-party provider.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from Mixpanel sources that could contain malicious instructions designed to manipulate the agent's behavior.
Ingestion points: Mixpanel event data from MIXPANEL_AGGREGATE_EVENT_COUNTS and user profiles from MIXPANEL_QUERY_PROFILES.
Boundary markers: No explicit instructions or delimiters are used to separate untrusted data from the agent's internal prompts.
Capability inventory: The skill possesses the ability to modify data via MIXPANEL_PROFILE_BATCH_UPDATE and execute remote scripts via MIXPANEL_JQL_QUERY.
Sanitization: The skill does not define methods for sanitizing or validating ingested Mixpanel data before it is processed.

mixpanel-automation