Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
Hosted MCP servers for Salesforce and AI clients
This repo is the entry point for Salesforce’s hosted MCP servers. It explains how AI assistants connect to Salesforce through MCP and where to find setup, server reference, and community guidance.
Builders who want their MCP client to connect securely to Salesforce.
You can let an AI assistant reach Salesforce tools and assets through one standard connection instead of custom integrations.
What it does
Hosted MCP server access
Connects MCP-compatible clients to Salesforce through a single interface.
Enterprise authentication and governance
Uses Salesforce authentication and governance controls for these connections.
Developer reference
Links to the hosted MCP server guide, including available servers, tools, and authentication details.
Community documentation hub
Uses the wiki and GitHub issues for patterns, questions, and feature ideas.
README
Salesforce Hosted MCP Servers
Salesforce’s hosted MCP servers allow AI assistants like Claude, ChatGPT, and more connect securely to your Salesforce logic and assets using the Model Context Protocol. Hosted MCP servers are now generally available (GA).
MCP is an open standard — think of it as USB-C for AI integrations. Any MCP-compatible client can connect to Salesforce through a single, standardized interface with enterprise-grade authentication and governance built in.
Documentation
Documentation for MCP servers is available in multiple locations, each serving a different audience and purpose. Here, we use a car analogy to explain each of the distinct documentation options.
![]() |
Basic Administration — Salesforce Help & TrainingConfigure MCP Servers in Setup Enable and disable servers, create custom servers, attach APIs and Flows, manage External Client Apps. Help & Training covers questions about anything you'd do in Salesforce Setup. Help & Training teaches you the basics of how to drive this new car. |
![]() |
Server Reference and Building Custom Servers — Salesforce DevelopersDeveloper Guide for Hosted MCP Servers Browse available servers and tools, build custom MCP servers, and understand authentication for Claude, ChatGPT, et. al. If your question is about which APIs are available or how to authenticate, start here. Developer docs represent the owner's manual with detailed specs. |
![]() |
Best Practice Patterns & Community — This WikiThis repo will become a hub for more advanced development and architecture topics, updated on an ad-hoc basis as the MCP space evolves. Best practices, architecture patterns, client-specific concerns, and more can change on a weekly basis. This wiki provides a channel for all of us to keep each other up-to-date. The wiki and issues list are a mechanic's forum where people share what works in the real world. |
Quick Links
Have a Question or Idea?
At this early stage, we are offering GitHub Issues to track feedback on the MCP server experience:
- Suggest a doc improvement — Flag gaps, unclear instructions, or missing examples
- Request a feature — Suggest new tools, servers, or capabilities
- Ask a question — Anything else about MCP servers
The intent here is to foster richer discussions with our developer community. For connection problems or potential bugs, please contact Salesforce Support.
Note: There is no SLA associated with issues in this repo. For support from Salesforce, you must use the official channels. Issues that otherwise would be support cases will be closed.
Files in the repo
- .github
- documentation
- .gitignore
- CODE_OF_CONDUCT.md
- CODEOWNERS
- CONTRIBUTING.md
- LICENSE.txt
- README.md
- SECURITY.md
Discussion (0)
Ask about usage, or say what you built with itSign in to join the discussion.
No comments yet. Be the first to say what this is good for.
More connectors
High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama…) with Codex CLI, App, SDK, and Claude Code
Git-native persistent memory for AI coding agents. Implements Google OKF v0.2 with sub-300µs in-memory BM25 search, embedded MCP server, and progressive disclosure. Slashes token bloat by 80% with zero external databases or dependencies. Built in pure Go.
Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.
Stop your AI from making things up — it proposes, deterministic tools decide, every claim checked against ground truth with evidence. Grounded facts and context survive resets. Reverse engineering is the proving ground. MCP server + CLI.


