# LaunchDarkly MCP connector

The LaunchDarkly connector brings feature flags, AgentControl configs and observability data into Claude, ChatGPT, Cursor and the Elaichi Agent, so each person works with LaunchDarkly under their own access and every change is logged.

Source: https://elaichi.ai/connectors/launchdarkly/

## Facts

| | |
| --- | --- |
| Application | LaunchDarkly |
| Category | Feature Flag |
| Authentication | Connects over OAuth |
| Bring your own OAuth app | No |
| Native MCP | Yes. LaunchDarkly builds and runs this MCP server. Elaichi adds sign-in, access controls and an audit log on top |
| Support for its tools | support.launchdarkly.com/hc/en-us |
| MCP endpoint | https://api.elaichi.ai/mcp |
| Works with | Claude, ChatGPT, Cursor, any MCP client, and the Elaichi Agent |
| Tools advertised by name | No. Connected tools are never listed one by one, however few there are. The endpoint advertises `search_tools` and `execute_tool` instead |

## What you can ask once LaunchDarkly is connected

- Which flags are on in production but off in staging?
- Turn on the new checkout flag for the beta segment.
- Show error rates since the pricing page flag went live.

## Connect LaunchDarkly in Elaichi

This happens once for the organization, before any client is involved.

1. Open Connections, choose Add connection, and pick LaunchDarkly.
2. Optionally set Share with, then press Connect.
3. Approve it in LaunchDarkly. LaunchDarkly's own window opens. Whoever approves it decides what this connection can reach.

Credentials are vaulted and nobody, including the AI, reads them back. The connection becomes a toolbox immediately, so you can curate which LaunchDarkly tools are exposed, rename them, or freeze arguments before anyone points a client at it.

## LaunchDarkly MCP connector for Claude

Endpoint: https://api.elaichi.ai/mcp

1. Open Customize, then Connectors.
2. Press Add.
3. Name it, paste the MCP server URL, then Continue.
4. Sign in and approve.

On Team and Enterprise, an Owner adds it once. Everyone else turns it on for themselves.

## LaunchDarkly MCP connector for ChatGPT

Endpoint: https://api.elaichi.ai/mcp

1. Open Plugins, then press the + button.
2. Name it and paste the endpoint into Server URL.
3. Leave Authentication on OAuth, then tick the risk acknowledgement.
4. Press Create, then sign in and approve.

Works on the web today. The plugin directory lives at chatgpt.com/plugins.

## LaunchDarkly MCP connector for Cursor

Endpoint: https://api.elaichi.ai/mcp

1. Open `~/.cursor/mcp.json`.
2. Add the endpoint under `mcpServers`.
3. Reload Cursor, then sign in and approve.

Set up per machine, so repeat it on each computer you work from.

## Connect LaunchDarkly to any MCP client

Endpoint: https://api.elaichi.ai/mcp

1. Add the endpoint as a remote MCP server.
2. Sign in and approve.

The Elaichi Agent already has these tools, with nothing to set up.

## What the consent screen decides

Only Read is granted by default, which is not enough to call a LaunchDarkly tool. Over MCP there is no trusted place to confirm a write in the moment, so the consent screen is the standing approval rather than a formality. Grant Read and Run tools. Think hard before granting Delete, which reaches into connected apps and cannot be undone.

## What teams do with LaunchDarkly through Elaichi

### Check what is live before a deploy

Engineering. Ask which flags are on in production but still off in staging, and who changed them last, before shipping anything that depends on them.

### Roll a feature out to one segment

Product. Turn a flag on for the beta segment or a single customer in LaunchDarkly, then widen the rollout once feedback comes back.

### Find stale flags ready for cleanup

Release. List flags that have been fully on or fully off for months across every environment, and hand engineering a cleanup list.

### Confirm what a customer can see

Support. Check whether a customer is targeted by a flag before replying to a ticket, instead of asking an engineer to look it up.

### Tune an AgentControl config in place

AI platform. Review an AgentControl config in LaunchDarkly, adjust the model or prompt variation, and record why it changed.

### Read observability data after a toggle

SRE. Pull error and session data from LaunchDarkly for the hour after a flag went on, and decide whether to roll it back.

## Elaichi vs Zapier MCP vs Composio for LaunchDarkly

All three can connect LaunchDarkly to an AI assistant, and all three have admin controls. They differ in where access lives and how you pay.

| What to check | Elaichi | Zapier MCP | Composio |
| --- | --- | --- | --- |
| Where the AI connects | One address for the whole organization. Endpoint: https://api.elaichi.ai/mcp | A server per member, created at sign-in. | An MCP endpoint per team, or an SDK. |
| Control over LaunchDarkly tools | Allow or restrict single LaunchDarkly tools, per role or user. | App and action restrictions on the account. | Role permissions, down to the action. |
| Record of calls | One audit entry per LaunchDarkly call. | A History tab of tool calls. | A log of every tool call. |
| Single sign-on | SAML or OIDC, plus SCIM, on Gold. | SAML on Enterprise. | SAML and OIDC on Enterprise. |
| Price | $15 per user per month. | 2 tasks per successful call. | Billed per tool call. |

Sources: Zapier MCP [docs](https://docs.zapier.com/mcp/get-started/quickstart), [security](https://docs.zapier.com/mcp/manage/security), [usage](https://docs.zapier.com/mcp/features/usage); Composio [docs](https://docs.composio.dev/docs/composio-connect), [gateway](https://composio.dev/mcp-gateway), [enterprise](https://composio.dev/enterprise), [pricing](https://composio.dev/pricing). Checked September 2026.

Longer take: [Zapier MCP alternative](/blog/zapier-mcp-alternative/) and [when you don't need an MCP gateway](/blog/when-you-dont-need-an-mcp-gateway/).

## Frequently asked questions

### How do I connect LaunchDarkly to Claude?

First connect LaunchDarkly in Elaichi, where you sign in to your LaunchDarkly account over OAuth and approve the access, with no OAuth application to register and no client ID or secret to generate. Then open Claude, go to Customize, then Connectors, then Add, and paste https://api.elaichi.ai/mcp. Claude asks you to sign in to Elaichi, and from then on it works with LaunchDarkly as you.

### Does LaunchDarkly work with ChatGPT and Cursor as well as Claude?

Yes. Once LaunchDarkly is connected in Elaichi, the same endpoint, https://api.elaichi.ai/mcp, works in Claude, ChatGPT, Cursor, any other MCP client and the Elaichi Agent. You connect LaunchDarkly once and every client picks it up.

### What can an AI agent actually do with my LaunchDarkly data?

It can tell you which feature flags are on in each environment, turn a flag on or off for a segment, review and adjust AgentControl configs, and read observability data such as errors after a rollout. What it can reach depends on what your LaunchDarkly account has set up. Short concrete asks, like naming the flag and the environment, get better results than long sentences.

### Does connecting LaunchDarkly give the AI access to every project and environment?

No. Access follows the person who signed in, so the AI sees only the LaunchDarkly projects, environments and flags that person already has permission to see. Elaichi can narrow that further, for example to read-only or to one environment, but it can never widen what LaunchDarkly itself allows.

### Can my team share one LaunchDarkly connection?

Yes. One person connects LaunchDarkly in Elaichi and shares it with a team, and nobody else ever handles a token or credential. Each teammate still signs in to Elaichi as themselves, so every flag change in the audit log names the person who made it.

### Can I stop an agent from deleting or changing things in LaunchDarkly?

Yes. In Elaichi, restrictions on the LaunchDarkly connection work per action, so you can allow reading flags while blocking changes, deletions or edits to AgentControl configs. A restricted action is never advertised to Claude, ChatGPT or Cursor, so no prompt can reach it.

### What happens to a LaunchDarkly connection when someone leaves?

Offboarding that person in Elaichi ends their access to LaunchDarkly through every client at once. If they shared a LaunchDarkly connection with a team, it keeps working for everyone else. Disconnecting LaunchDarkly once in Elaichi removes it from Claude, ChatGPT, Cursor and every other client.

### Does the LaunchDarkly MCP connector work with Gemini, Codex, Claude Code or other MCP clients?

Yes. LaunchDarkly is reached over the same MCP endpoint every client uses, so anything that speaks MCP can call it — Gemini, Codex, Claude Code, Windsurf, Cline, Zed and OpenCode among them — alongside Claude, ChatGPT, Cursor, and the Elaichi Agent. The tools on offer and the access behind them are identical whichever client asks. Only the setup screen differs.

### Is Elaichi an alternative to Zapier MCP for LaunchDarkly?

Yes. Both let Claude, ChatGPT or Cursor use LaunchDarkly. Zapier MCP fits a team that already automates in Zapier, since each person signs in and acts as themselves in that account. Elaichi fits when IT wants one address for the whole company, per-tool rules by role, and a record of every LaunchDarkly call.

### How is Elaichi different from Composio for LaunchDarkly?

Composio gives AI agents tools and sign-in handling across 1,000+ apps, for developers building agents or people using an assistant, billed per tool call. Elaichi gives a company's own people governed access to LaunchDarkly: one address, restrictions per role or user, and $15 per user per month. Both have role permissions and a log of every call.