Receber por email
← Voltar para o catálogo

claude-skills-vault

A curated collection of high impact skills for Claude Code designed to supercharge the senior full stack workflow. This vault automates the repetitive parts of development like architectural reviews, TDD cycles, and PR management so you can stay in flow. It is a force multiplier for shipping clean, production ready code at scale. 🚀⚡️

25estrelas
Atualizado há 4 meses

Ver no GitHub ↗ Licença: MIT

Como adicionar

/plugin marketplace add georgekhananaev/claude-skills-vault

O comando exato pode variar conforme o repositório. Confira o README no GitHub.

Para o autor da skill

Cole no README do seu repo

Mostra que sua skill está catalogada na Skillteca, gera backlink e tráfego rastreável.

[![Skillteca](https://www.skillteca.com.br/api/badge/claude-skills-vault-mpsybo6z/svg)](https://www.skillteca.com.br/skills/claude-skills-vault-mpsybo6z)

Alerta por categoria

Receba novas skills de Design e Frontend toda segunda

1 email curto, só com as skills novas de Design e Frontend. 4 minutos de leitura, sem spam, cancela com 1 clique.

Você confirma o email no primeiro envio. Sem spam. Cancela com 1 clique.

Ler descrição completa

Monday.com Workspace Manager

Manage Monday.com workspaces through the official MCP server — create items, update statuses, move tasks between groups, archive/delete items, manage boards & columns, and add comments.

When to Use

Invoke when:

  • Creating, updating, or deleting items/tasks on Monday.com
  • Managing boards, groups, or columns
  • Moving items between groups or boards
  • Adding comments/updates to items
  • Querying board structure, items, or users
  • Setting up Monday.com MCP integration
  • Any Monday.com workspace automation

Prerequisites & Setup

First-Time Setup Flow

On first use, check if Monday.com MCP is configured. If not, guide the user through setup using AskUserQuestion:

Question: "How would you like to connect to Monday.com?"
Options:
  - "Hosted MCP (Recommended)" — No local install, auto-updates, OAuth auth
  - "Local MCP via npx" — Requires Node.js 20+, API token
  - "Direct API (curl/scripts)" — Manual GraphQL calls, no MCP dependency

Option A: Hosted MCP (Recommended)

Add to Claude Code MCP settings (~/.claude/settings.json or project .mcp.json):

{
  "mcpServers": {
    "monday-mcp": {
      "url": "https://mcp.monday.com/mcp"
    }
  }
}

OAuth handles auth automatically — no API token needed.

Option B: Local MCP via npx

  1. Get API Token: Monday.com → Avatar → Developers → My access tokens
  2. Configure MCP in Claude Code settings:
{
  "mcpServers": {
    "monday-api-mcp": {
      "command": "npx",
      "args": ["@mondaydotcomorg/monday-api-mcp@latest"],
      "env": {
        "MONDAY_TOKEN": "<your_token>"
      }
    }
  }
}

Useful flags:

| Flag | Values | Purpose | |------|--------|---------| | --read-only / -ro | true | Read-only mode | | --enable-dynamic-api-tools / -edat | true / only | Full GraphQL access (beta) | | --version / -v | 2025-07 | API version |

Option C: Direct API (No MCP)

For environments w/o MCP support. Set MONDAY_API_TOKEN in .env or .env.local at project root:

MONDAY_API_TOKEN=your_token_here

The script auto-loads from .env.local (priority) or .env. Alternatively, export the env var directly.

Use scripts/monday_api.sh for direct GraphQL calls.

Tool Selection: MCP vs API Script

On every Monday.com request, determine which tool to use:

1. Check: Are Monday.com MCP tools available? (ToolSearch "monday")
   → YES: Use MCP tools — go to decision table below
   → NO:  Go to step 2

2. Check: Does .env or .env.local have MONDAY_API_TOKEN?
   → YES: Use scripts/monday_api.sh with GraphQL queries
   → NO:  Ask user to set up (AskUserQuestion with setup options above)

MCP vs API Decision Table

| Operation | Best Tool | Why | |-----------|-----------|-----| | Single item CRUD (create, update, delete) | MCP | One tool call, handles auth, structured params | | Get board schema/info | MCP (get_board_info) | Returns columns, groups, filtering guidelines in one call | | Get items w/ filters | MCP (get_board_items_page) | Built-in filtering, ordering, pagination, includeColumns toggle | | Search items by text | MCP (get_board_items_page w/ searchTerm) | Fuzzy search built-in | | Assign owner | MCP (change_item_column_values) | Single call w/ personsAndTeams JSON | | List users/teams | MCP (list_users_and_teams) | Supports getMe, name search, team members | | Create board/group/column | MCP | Type-safe params, enum validation | | Get column type info | MCP (get_column_type_info) | Returns JSON schema for column settings | | Board activity logs | MCP (get_board_activity) | Date range filtering built-in | | Move board/folder | MCP (move_object) | Handles positioning | | Workspace management | MCP | Create, update, list workspaces | | Forms | MCP | Create and get forms | | Multi-board dashboard | API script | MCP queries one board at a time; script queries all boards in one GraphQL call | | Batch updates (5+ items) | API script | One GraphQL mutation w/ aliases vs 5+ separate MCP calls | | Batch deletes | API script | Same — aliases batch multiple deletes into one call | | Webhooks | API script | MCP has no webhook tools | | Archive item/board | API script | MCP has no archive mutation | | Duplicate item | MCP (create_item w/ duplicateFromItemId) | Built-in duplicate support | | Subitems | MCP (create_item w/ parentItemId) | Native subitem creation |

Performance Rules

  1. Single operations → always MCP (faster, no shell overhead, structured errors)
  2. Bulk operations (5+ items) → API script w/ aliases (1 HTTP call vs N MCP calls)
  3. Cross-board reads → API script (query multiple boards in one request)
  4. Filtered reads → MCP (built-in filter operators: any_of, between, contains_text, greater_than, etc.)
  5. MCP unavailable → API script for everything

Safety Classification

| Tier | Operations | Behavior | |------|-----------|----------| | Safe | List boards, get items, get schema, list users | Execute immediately | | Write | Create item, update columns, add comment, create group/column | Inform user → execute | | Destructive | Delete item, delete column, archive board | AskUserQuestion → confirm → execute |

Decision Flow

Request received
  → Classify tier (see table above)
  → Safe?        Execute immediately, show results
  → Write?       Show what will change → execute
  → Destructive? AskUserQuestion w/ confirm/cancel → execute or abort

Destructive Operation Confirmation

For delete/archive operations, always confirm:

Question: "Delete item 'Task Name' (ID: 123456) from board 'Project Board'?"
Options:
  - "Delete permanently" — Item cannot be recovered
  - "Cancel" — Keep item unchanged

Available MCP Tools

Item Operations

| Tool | Tier | Description | |------|------|-------------| | create_item | Write | Create item w/ column values | | change_item_column_values | Write | Update item columns | | move_item_to_group | Write | Move item between groups | | delete_item | Destructive | Permanently delete item | | get_board_items_by_name | Safe | Search items by name | | create_update | Write | Add comment to item |

Board Operations

| Tool | Tier | Description | |------|------|-------------| | create_board | Write | Create new board | | get_board_schema | Safe | Get columns, groups, structure | | create_group | Write | Add group to board | | create_column | Write | Add column to board | | delete_column | Destructive | Remove column from board |

Account Operations

| Tool | Tier | Description | |------|------|-------------| | list_users_and_teams | Safe | Get workspace users & teams |

Dynamic API Tools (Beta)

Enable w/ --enable-dynamic-api-tools true for full GraphQL access:

| Tool | Description | |------|-------------| | all_monday_api | Execute any GraphQL query/mutation | | get_graphql_schema | Explore API schema | | get_type_details | View type information |

Common Workflows

CRITICAL: Always Discover Board Schema First

Column IDs are unique per board (e.g. status vs project_status, person vs project_owner). Never assume column IDs — always query the board schema first.

Step 0 (always): Get board schema to discover column IDs, group IDs, and user IDs
  → query: { boards(ids: [BOARD_ID]) { columns { id title type } groups { id title } } }
  → query: { users { id name email } }

Create Item w/ Owner & Status

1. Get board schema → find column IDs for people, status, date, etc.
2. Get user ID → query { users { id name } } or { me { id } }
3. Create item using discovered column IDs:

``

Compartilhar X LinkedIn