monday-com_skill
- Python
9
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill georgekhananaev/claude-skills-vault --skill monday-com- SKILL.md13.3 KB
Overview
This skill manages Monday.com workspaces through the official MCP server or a direct API fallback. It streamlines creating, updating, moving, archiving, and deleting items, boards, groups, columns, and comments while guiding interactive setup for hosted or local MCP deployments. Use it to query workspace structure, perform safe writes, and run bulk operations with the correct tool chosen automatically.
How this skill works
On each request the skill detects available tooling: hosted MCP, a local MCP via npx, or a direct GraphQL API script using MONDAY_API_TOKEN. It inspects board schema first (columns, groups, users) to avoid incorrect column IDs, classifies operations by safety tier (safe, write, destructive), and routes work to MCP tools for single-item operations or to a GraphQL script for batch and cross-board queries. Destructive actions always prompt for explicit confirmation.
When to use it
- Create, update, move, archive, or delete items and boards on Monday.com
- Add comments/updates to items or change column values
- Discover board schema, column types, groups, or user lists before edits
- Run batch updates or cross-board reads (use API script for efficiency)
- Set up or reconfigure MCP integration (hosted or local)
Best practices
- Always query board schema first to get correct column and group IDs
- Use MCP for single CRUD operations; use API script for bulk (5+ items) or multi-board queries
- Classify operations by tier: execute safe reads immediately, show planned changes for writes, require confirmation for destructive actions
- Prefer hosted MCP for simplest auth; use local MCP when you need env-controlled tokens or --enable-dynamic-api-tools
- Use GraphQL aliases in batch mutations to reduce HTTP calls and stay within rate limits
Example use cases
- Create a task with owner and status: discover schema → get user ID → create_item with personsAndTeams JSON
- Move multiple tasks into a release group: get target group ID → move_item_to_group for each item (or batch via API script)
- Generate a workspace dashboard: query all boards and items in a single GraphQL call, group by status and owner
- Archive/delete a board or column: run confirm prompt (destructive) then call archive/delete via API script if MCP lacks archive support
- Add a comment/update to an item using create_update MCP tool
FAQ
It prefers MCP tools for single-item CRUD because they handle auth and structured params with minimal overhead.
When should I use the direct API script instead of MCP?
Use the API script for bulk operations (5+ items), cross-board queries, webhooks, or when MCP is unavailable.