The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external Wrike entities such as task descriptions and folder titles which could contain malicious instructions. * Ingestion points: The skill ingests data through WRIKE_FETCH_ALL_TASKS, WRIKE_GET_TASK_BY_ID, and WRIKE_GET_FOLDERS (SKILL.md). * Capability inventory: The skill has access to high-impact tools including WRIKE_DELETE_FOLDER and WRIKE_DELETE_SPACE (SKILL.md). * Boundary markers: No delimiters or specific instructions are provided to separate retrieved data from system prompts. * Sanitization: No sanitization or validation of external content is mentioned.
[EXTERNAL_DOWNLOADS]: The setup instructions require the AI agent to connect to an external MCP server located at https://rube.app/mcp.
[COMMAND_EXECUTION]: The skill provides tools for permanent and irreversible deletion of folders, projects, and spaces within the Wrike platform.

wrike-automation