mcp s

This MCP server collection provides a set of modular servers that extend Claude's capabilities with real-world integrations. Each server exposes a standard interface so Claude can perform tasks such as web searching (Brave Search), retrieving Brazilian public data (Brasil API), managing Google Calendar events, handling CRM data (Ploomes CRM), and interacting with Slack workspaces (channels, messages, user profiles). To start a specific server, you can run it through NPX using the corresponding package name, for example: npx -y @integrabot/brave-search. Once running, Claude can call the server's endpoints through MCP to perform the supported actions. The collection is designed to be used together or individually, depending on the user's workflow and data access needs.

Prerequisites:

• Node.js (LTS) and npm installed on your machine
• Git (optional, for cloning the repository)
• Access tokens or API keys as required by individual services (see environment variables)

Installation steps:

1. Clone or download the repository containing the MCP Servers Collection.
2. Install dependencies at the repository root:
   • npm install
3. Build the collection (if the repository provides a build step):
   • npm run build
4. (Optional) Run in development mode to watch for changes:
   • npm run watch
5. Start individual MCP servers using NPX (as shown in the mcp_config):
   • npx -y @integrabot/brave-search
   • npx -y @integrabot/brasil-api
   • npx -y @integrabot/google-calendar
   • npx -y @integrabot/ploomes-crm
   • npx -y @integrabot/slack

Note: If you need to customize environment variables for a server, set them in your environment before starting the server. Examples are provided in the additional_notes section.

Environment variables and configuration hints:

• Brave Search: BRAVE_SEARCH_API_KEY or relevant credentials expected by the Brave Search integration.
• Brasil API: BRASIL_API_KEY or required authentication details; may also require endpoints like CEP or CNPJ configurations.
• Google Calendar: GOOGLE_CALENDAR_OAUTH_TOKEN or API key; for OAuth2 flows you may need client ID/secret and redirect URIs configured.
• Ploomes CRM: PLOOMES_API_KEY or OAuth credentials as required by the Ploomes integration.
• Slack: SLACK_BOT_TOKEN and SLACK_APP_TOKEN for bot interactions; may also require a default workspace or channel configuration.

Common issues:

• API keys not being picked up: ensure environment variables are exported in the shell before launching the NPX command.
• Incorrect scope or OAuth misconfigurations: verify that the OAuth credentials have the correct scopes and are allowed for the target project.
• Network restrictions: some endpoints may be blocked by corporate networks; use VPN or allowlist required domains if applicable.
• Version mismatches: keep the MCP SDK and the individual packages up to date to avoid protocol incompatibilities.

Configuration tips:

• When running multiple servers locally, consider using separate ports or containerized environments to avoid port collisions.
• If you plan to publish or share this collection, ensure you have proper API keys secured (do not commit secrets to version control).