sf-intelligence — Salesforce org intelligence for AI agents
Offline read-only MCP server that maps one Salesforce org's metadata and impact
- Pricing
- Open source
- Platforms
- CLI, self-hosted
- Runs
- Locally
- Version
- 0.3.3
- Listing updated
- 1 day ago
About sf-intelligence — Salesforce org intelligence for AI agents
sf-intelligence — Salesforce org intelligence for AI agents is an open-source MCP server for sales and crm. It runs locally on your machine and is started by your AI client.
sf-intelligence answers from a local folder holding a snapshot of one org's metadata and a dependency graph, so it works offline and never writes to Salesforce. Every answer names the components behind it and lists what was not checked, for example reports or Lightning pages missing from the snapshot, rather than implying nothing is there.
It targets admin and developer questions: where a field is used, whether it can be deleted, what runs in what order when a record is saved, which Apex is unused and who can see or edit an object. The README includes a typo-tolerant resolver that asks a clarifying question when a name is ambiguous, and a built-in demo org for trying it without a Salesforce login.
By default the assistant sees 16 tools; further analyses sit behind a gateway tool. An optional live read-only mode is off until switched on. The vault location is set with the SFI_VAULT environment variable.
Install
Run in a terminal
claude mcp add sf-intelligence --env SFI_VAULT='<SFI_VAULT>' -- npx -y sf-intelligenceAdd to claude_desktop_config.json
{
"mcpServers": {
"sf-intelligence": {
"command": "npx",
"args": [
"-y",
"sf-intelligence"
],
"env": {
"SFI_VAULT": "<SFI_VAULT>"
}
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"sf-intelligence": {
"command": "npx",
"args": [
"-y",
"sf-intelligence"
],
"env": {
"SFI_VAULT": "<SFI_VAULT>"
}
}
}
}Add to .vscode/mcp.json
{
"servers": {
"sf-intelligence": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"sf-intelligence"
],
"env": {
"SFI_VAULT": "<SFI_VAULT>"
}
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"sf-intelligence": {
"command": "npx",
"args": [
"-y",
"sf-intelligence"
],
"env": {
"SFI_VAULT": "<SFI_VAULT>"
}
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"sf-intelligence": {
"command": "npx",
"args": [
"-y",
"sf-intelligence"
],
"env": {
"SFI_VAULT": "<SFI_VAULT>"
}
}
}
}Configuration
SFI_VAULTAbsolute path to the org-kb vault this server should answer from, e.g. <your-sfdx-project>/org-kb. Set this: without it the server falls back to ./org-kb relative to whatever directory your MCP host happens to launch in, and an MCP host is almost never launched inside your Salesforce project. The server still starts without it and reports what it looked for via sfi.setup_status, but it will have no org to answer about.Required
Tools this server provides (8)
Taken from the project documentation; the running server may differ.
- sfi.resolve
- Resolves messy phrasing to a component
- sfi.search_components
- Searches org components
- sfi.get_component
- Fetches one component
- sfi.health_check
- Checks the server's state
- sfi.list_analyses
- Lists available analyses
- sfi.describe_analysis
- Describes an analysis
- sfi.run_analysis
- Gateway tool that runs other analyses
- sfi.live_consent
- Handles consent for the optional live read-only mode
What it does
- Field usage lookup
- Finds Apex, flows, layouts, rules and reports that reference a field.
- Delete verdicts
- Gives safe, risky, blocking or review with reasons and what was not checked.
- Save-order trace
- Lists triggers, validations, flows and approvals in run order.
- Dead Apex detection
- Finds classes and triggers nothing calls, with confidence tiers.
- Permission analysis
- Explains who can see or edit via profiles and permission sets.
- Built-in demo org
- Lets you try it with no Salesforce login.
Good for
- Check what breaks before deleting a Salesforce field
- Trace what runs when a record is saved
- Find unused Apex before a cleanup
- Understand why a user cannot see a record
Pricing
Open source
Described as free; runs locally
Prices change. Check sfi.auditforce.cloud before you buy.
sf-intelligence — Salesforce org intelligence for AI agents FAQ
Does it change my org?
No, it is read-only and runs on your computer.
Does it need a network connection?
Answers are served offline from a local vault; the live read-only plane is optional and off by default.
How do I try it without an org?
Run the demo with npx and sf-intelligence demo.
What must I configure?
SFI_VAULT, the absolute path to your org-kb folder.
Is sf-intelligence — Salesforce org intelligence for AI agents free?
Yes. sf-intelligence — Salesforce org intelligence for AI agents is open source, so you can use and self-host it for free.
How do I add sf-intelligence — Salesforce org intelligence for AI agents to Claude Code?
Run this command in your terminal: claude mcp add sf-intelligence --env SFI_VAULT='<SFI_VAULT>' -- npx -y sf-intelligence
What are the best alternatives to sf-intelligence — Salesforce org intelligence for AI agents?
The closest alternatives to sf-intelligence — Salesforce org intelligence for AI agents are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.
Alternatives to sf-intelligence — Salesforce org intelligence for AI agents

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Codebase Memory
MCP serverLocal MCP server that builds a code knowledge graph for AI coding agents
Runs locally- Open source
- 48.7K GitHub stars
- Knowledge and memory

GitHub
MCP serverOfficial MCP server connecting AI assistants to GitHub repos, issues and workflows
Not checked yet- Remote
- Open source
- 34.5K GitHub stars
- Developer tools

Serena MCP: the IDE for your agent
MCP serverMCP toolkit giving coding agents symbol-level code retrieval and editing
Runs locally- Open source
- 31.5K GitHub stars
- Developer tools

XERJ
MCP serverLocal search engine for AI agents with cited answers over your code and docs
Runs locally- Open source
- 19.2K GitHub stars
- Knowledge and memory

Ouroboros
MCP serverSpec-first layer that pins requirements and verifies results of AI coding agent work
Runs locally- Open source
- 6.4K GitHub stars
- Developer tools