Backlot
Run Slack, Gmail, Drive, GitHub and Jira APIs locally over one test corpus
- Pricing
- Open source
- Platforms
- CLI, self-hosted, API
- Runs
- Locally
- Version
- 0.1.1
- Listing updated
- 2 days ago
About Backlot
Backlot is an open-source MCP server for communication. It runs locally on your machine and is started by your AI client.
Backlot is aimed at developers who need the behaviour of an enterprise API, not only a canned response. It serves vendor-shaped endpoints, including pagination, authentication, errors and per-document access controls, from one SQLite corpus, so official SDKs work after changing a single base URL.
After pip install, backlot import --bundled loads a sample corpus and backlot serve exposes every supported API on one local port. You can also import your own JSONL corpus, one source document per line, and every user in it gets deterministic tokens so you can check that one caller never sees another's documents. It emulates a documented subset of each API, listed in an endpoint matrix, rather than every vendor endpoint.
Running backlot mcp serves the sources to an agent as MCP tools over stdio, optionally answering as a given user. Write APIs exist for some vendors, so an agent that acts can be graded on what it did and the corpus reset between runs.
Install
Run in a terminal
claude mcp add backlot -- uvx backlotAdd to claude_desktop_config.json
{
"mcpServers": {
"backlot": {
"command": "uvx",
"args": [
"backlot"
]
}
}
}Add to ~/.cursor/mcp.json
{
"mcpServers": {
"backlot": {
"command": "uvx",
"args": [
"backlot"
]
}
}
}Add to .vscode/mcp.json
{
"servers": {
"backlot": {
"type": "stdio",
"command": "uvx",
"args": [
"backlot"
]
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"backlot": {
"command": "uvx",
"args": [
"backlot"
]
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"backlot": {
"command": "uvx",
"args": [
"backlot"
]
}
}
}What it does
- Vendor-shaped responses
- Reproduces pagination, auth, errors and response shapes.
- Per-document ACLs
- Generated users, groups and tokens enforce who sees what.
- Many services in one process
- Slack, Gmail, Drive, Teams, GitHub, Jira, Confluence, Notion, Linear, HubSpot, Fireflies and S3.
- Your own corpus
- Import a JSONL file and validate it with a dry run.
- MCP mode
- backlot mcp exposes every source to an agent over stdio.
- Write overlay
- Lets agents that act be graded, then reset the corpus.
Good for
- Test an integration without a vendor account or OAuth approval
- Run CI with no secrets or network
- Evaluate a RAG pipeline against fixed data
- Grade an agent's actions against a resettable workspace
- Reproduce a parser bug with a shaped test document
Pricing
Open source
Open-source Python package on PyPI
Backlot FAQ
Do I need vendor accounts?
No. Backlot supplies the data and credentials locally.
Does it cover every endpoint?
No, only the documented subset listed in its supported-sources matrix.
How do I install it?
pip install backlot, then backlot import --bundled and backlot serve.
Can agents use it?
Yes, backlot mcp serves the sources as MCP tools over stdio.
Is Backlot free?
Yes. Backlot is open source, so you can use and self-host it for free.
How do I add Backlot to Claude Code?
Run this command in your terminal: claude mcp add backlot -- uvx backlot
What are the best alternatives to Backlot?
The closest alternatives to Backlot are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.
Alternatives to Backlot

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