dingtalk-ai-table

What This Skill Does

The dingtalk-ai-table skill enables OpenClaw AI agents to interact directly with DingTalk's advanced AI Tables (Multidimensional Tables). By leveraging the official new-version MCP (Model Context Protocol) schema, this skill allows for granular management of bases, tables, fields, and records using standardized identifiers (baseId, tableId, fieldId, recordId). It streamlines complex data operations, from creating new collaborative tables and defining custom schema structures to executing bulk CRUD operations on records. The skill includes a robust automated version-gate mechanism that ensures the underlying MCP server is compatible with the new schema, preventing errors caused by legacy configuration.

Installation

1. Ensure you have the mcporter CLI installed: npm install -g mcporter.
2. Get your unique Streamable HTTP URL from the DingTalk MCP Market.
3. Register the server in your environment or via CLI: mcporter config add dingtalk-ai-table --url "<YOUR_URL>".
4. Install the skill into OpenClaw: clawhub install openclaw/skills/skills/aliramw/dingtalk-ai-table.
5. Upon first execution, the system will verify the server schema and cache the validation in ~/.openclaw/workspace/.cache/dingtalk-ai-table/ to ensure performance efficiency.

Use Cases

• Project Management: Automatically create new bases and tables based on incoming project briefs or meeting notes.
• Data Synchronization: Synchronize local data logs or survey results into structured DingTalk AI tables via automated record insertion.
• Workflow Automation: Build dynamic field structures (e.g., adding status, owner, or date fields) programmatically to match changing business logic.
• Knowledge Management: Query existing base records to extract insights or populate internal dashboards using AI natural language processing.

Example Prompts

1. "Create a new table named 'Customer Feedback' in the CRM base with fields for 'Customer Name', 'Rating', and 'Comment'."
2. "Search for all records in the 'Project Tasks' table where the status is 'Pending' and update them to 'In Progress'."
3. "List all my available bases and tell me the structure of the table with the most recent modifications."

Tips & Limitations

• Version Guard: If you encounter a legacy schema error, strictly follow the prompt to update your URL from the DingTalk MCP Market. Do not attempt to bypass the schema check.
• Data Privacy: The MCP URL contains an authentication token; treat it with the same security as a password. Never share it in public logs or repositories.
• Scope: This skill is exclusively for the new DingTalk AI Tables MCP protocol. It does not support deprecated legacy DingTalk API patterns.