SALESFORCE ORG
INTELLIGENCE YOUR AI
CAN TRUST.

$ what is sf-intelligence?

Salesforce org intelligence for AI agents - grounded, read-only, offline.

A knowledge base for one org.

sf-intelligence is a free, source-available Model Context Protocol (MCP) server and CLI that turns one Salesforce org's metadata into a local, queryable knowledge base for AI agents like Claude and Cursor. You retrieve your org's metadata once; after that every answer is served offline and read-only from a vault on your machine - schema, fields, Apex, Flows, permissions, sharing, integrations, and OmniStudio - grounded in real metadata and labelled with its provenance. It never writes to your org and no org data leaves your machine. It is an independent project, not affiliated with Salesforce, Inc.

It runs as a read-only Salesforce MCP server for Claude, Cursor and Codex, pairs with portable Salesforce Agent Skills that interpret intent, and you can compare Salesforce org-intelligence tools to see where it fits.

the problem

A chatbot guesses. This reads the org.

Ask a general model “what breaks if I delete this field?” and it answers from generic Salesforce training data - confidently, and often wrong for your org. sf-intelligence answers only from the metadata you actually retrieved. When it isn't sure, it says so and shows you the candidates.

$ sfi.capabilities - what can you ask?

The questions your team actually asks.

Not a tool catalogue - the operational questions admins, developers, architects, release managers, and support reps ask every day. You type the question in plain language; the router surfaces the right tools and your AI assistant runs them - across eight areas.

› browse the full capability map› all 217 tools

use cases

Built for the questions that matter.

Intent-matched pages for the operational problems Salesforce teams search for - metadata analysis, dependency analysis, impact analysis, sharing troubleshooting, and MCP setup.

› all use cases › compare › agent-readiness › glossary

how it works

One retrieve, then ask anything.

No cloud service, no account. The whole index is files on your machine.

trust & provenance

Every answer is labelled.

Findings carry where they came from, how they were derived, and how complete the underlying coverage was - the full security and evidence model spells out every label.

Those three tiers grade a single edge - how one relationship was derived. An interpreted claim from sfi.interpret carries a second, separate claim confidence: the weakest of the firing rule's ceiling and the grounding edges it cites, so a claim is never asserted above the evidence beneath it - and with no citation there is no claim. glossary →

› how it's tested - ~9,300 tests, full CI gate & the trust model

before you point it at production

Safe to run, honest about limits.

The first question any admin asks. The left column is the one-screen guarantee sfi init prints in your terminal; the right is the boundaries it states plainly instead of papering over. More on AI safety →

highlights

Shipped for real orgs at scale.

Read-only capabilities built for large production orgs - all optional except the defaults you already get on refresh.

› full configuration reference

install

Register once, then just ask.

Distributed on npm as sf-intelligence - an MCP server plus the sfi CLI. Follow the guide to install sf-intelligence in a couple of minutes.

// Claude Code - project-scoped

claude mcp add --transport stdio --scope project sf-intelligence -- npx -y sf-intelligence mcp

// Claude Desktop or any MCP client

{
  "mcpServers": {
    "sf-intelligence": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "sf-intelligence", "mcp"]
    }
  }
}

Default is a 19-tool core roster (token-friendly, includes sfi.live_consent); every other analysis stays reachable via sfi.run_analysis. Set "env": { "SFI_TOOL_PROFILE": "full" } only if you want all 217 schemas advertised. details →

// first run, from your Salesforce DX repo

# 1. create the local vault and pick your org alias
sfi init
# 2. retrieve metadata and build the index (read-only)
sfi refresh --target-org my-org-alias
# 3. confirm freshness and component counts any time
sfi status

$ full getting-started guide