LaunchDarkly MCP Server
Connect to the LaunchDarkly MCP server to manage feature flags, control rollouts and targeting, manage segments, and check flag status using AI agents on Gumloop, Claude, or Cursor.
Try LaunchDarkly now
Type what you want done. Sign in and run it live with an AI agent.
|
LaunchDarkly
Installation
Get StartedCreate a Gumloop Account
To use this MCP, you need a Gumloop account. If you don't have one yet, you can create one for free.
Copy Your Server URL
Copy your MCP server URL and add it to your client. You'll be prompted to authorize on first use.
Tools (14)
List Projects
Retrieve all projects in your LaunchDarkly account. Projects contain feature flags and environments.
List Environments
Retrieve all environments within a project. Environments represent deployment targets like production, staging, test.
List Feature Flags
Retrieve all feature flags within a project with metadata and targeting rules.
Get Feature Flag
Retrieve a single feature flag by key with full configuration details.
Create Feature Flag
Create a new feature flag in a project.
Update Feature Flag
Update a feature flag using semantic patch instructions.
Delete Feature Flag
Delete a feature flag from a project.
Get Feature Flag Status
Get the status of a feature flag in an environment (new, active, inactive, launched).
List Code Repositories
List code repositories that have been connected for code references.
List Segments
List all segments in an environment. Segments are reusable groups of contexts that can be targeted by feature flags.
Get Segment
Get a single segment by key. Returns included/excluded contexts.
Create Segment
Create a new segment in an environment.
What is LaunchDarkly MCP?
The LaunchDarkly MCP server gives AI agents access to your LaunchDarkly feature management data. That means agents can list and read feature flags with their full targeting config, create flags (and clone targeting from an existing one), update flags through semantic-patch instructions like turning a flag on or off and adding targets or tags, delete flags, check flag lifecycle status per environment, manage reusable context segments, browse your projects and environments, and list the connected code repositories that reference your flags. It works across the core LaunchDarkly objects that engineering and product teams rely on: projects, environments, flags, and segments.
If your team spends time clicking through the LaunchDarkly dashboard to flip flags, edit targeting rules, audit which flags are still active, or stand up new segments by hand, an AI agent can take over a lot of that work. Describe what you need, and your AI agent will handle the flag and segment changes for you.
MCP stands for Model Context Protocol. It’s an open standard that gives AI agents a way to connect to external tools and services. Instead of writing code against the LaunchDarkly API, constructing low-level JSON patch diffs, and parsing responses yourself, you add your LaunchDarkly API access token to Gumloop once. After that, you can manage feature flags and segments just by chatting with your AI agent.
Related MCP servers
Github MCP Server
Server for interacting with Github
GitLab MCP Server
Interact with GitLab
Linear MCP Server
Interact with Linear issue tracking and project management
JIRA MCP Server
Interact with JIRA for project management, issue tracking, and user information access.
Datadog MCP Server
Interact with Datadog to monitor infrastructure, query metrics, search logs, and manage incidents
Slack MCP Server
Interact with Slack channels and messages
What you can do with LaunchDarkly MCP on Gumloop
Manage feature flags end to end
List flags in a project with filters, read a single flag with its full configuration and targeting, create new flags, and delete flags you no longer need. When you create a flag, an agent can clone targeting config from an existing flag to spin up a variant quickly.
Control rollouts and targeting with semantic patches
Update a flag through high-level semantic-patch instructions such as turning it on or off, adding or removing targets, renaming, and tagging. Your agent issues plain instructions instead of building raw JSON diffs, and it can apply changes even when scheduled changes conflict.
Preview changes before they go live
Run flag updates in dry-run mode to see exactly what would change without persisting anything. This makes a “plan then execute” pattern natural, so an agent can propose a rollout change and you can confirm before it commits.
Manage segments for audience targeting
List segments in an environment, read a segment with its included and excluded contexts, create new segments, update them by adding or removing contexts through semantic patches, and delete segments. Segments support arbitrary context kinds (user, business, device), and you can create unbounded segments for very large audiences.
Target by context kind, not just users
Because segments and flag targets operate on context kinds like user, business, and device, an agent can gate a feature for specific business accounts or devices, not just individual users.
Check flag health and lifecycle status
Pull a flag’s status in a given environment to see whether it’s new, active, inactive, or fully launched. An agent can roll this up across flags to flag stale toggles that are ready for cleanup.
Browse projects and environments
List every project in the account and every environment within a project (production, staging, test, and the rest). Projects and environments are read-only here, so an agent uses them to navigate and scope flag and segment work.
Track code references
List the connected code repositories tied to your flags so an agent can cross-reference where a flag is used in source and assess whether it’s still needed.
How to connect the Gumloop LaunchDarkly MCP Server
- 1
Create a free Gumloop account
Sign up at gumloop.com. No credit card required.
- 2
Add the LaunchDarkly MCP server
Copy your MCP server URL from Gumloop and add it to your preferred client (Claude, Cursor, or Gumloop workflows). You'll authorize on first use.
- 3
Start using LaunchDarkly in your AI workflows
That's it. Your AI agent can now manage feature flags, control rollouts and targeting, manage segments, and check flag status across your LaunchDarkly projects. Use it inside a Gumloop automation, in Claude Desktop, or in Cursor.
LaunchDarkly MCP use cases
Safe feature rollouts for platform engineers
When it’s time to ship a feature, a Gumloop agent can read the flag’s current targeting, run an update in dry-run mode to preview the change, then turn the flag on for a specific segment once you confirm. Engineers manage rollouts by chatting with the agent instead of clicking through targeting rules in the dashboard.
Flag cleanup and lifecycle audits for engineering managers
An agent can list every flag in a project, check each flag’s status per environment, and cross-reference the connected code repositories to spot flags that are inactive or no longer referenced in code. It compiles a cleanup candidate list into Google Sheets or Slack so the team can retire stale toggles with confidence.
Audience segment management for product teams
Product managers can ask an agent to create or update a segment that targets specific business accounts or device types, adding and removing contexts through semantic patches. The agent keeps reusable segments current so flag targeting stays accurate as customer lists change.
Incident kill-switches for on-call engineers
During an incident, an on-call engineer can tell an agent to turn off a problematic flag in production immediately, then read the flag status to confirm it’s no longer serving. Pair it with Slack so the agent posts what it changed and which environment it touched, keeping the response team in the loop.
Cross-platform flag operations
Connect LaunchDarkly with Slack, Jira, GitHub, and other MCP integrations in a single agent. When a Jira ticket moves to done, an agent can create the corresponding feature flag, clone targeting from a similar flag, and post the new flag key back to the ticket and a Slack channel, so flag setup tracks your engineering workflow automatically.
Why use Gumloop for LaunchDarkly MCP
Your API access token, stored securely
Most LaunchDarkly MCP servers you’ll find on GitHub make you store your access token in environment variables and write code to handle authentication and the API. With Gumloop you add your LaunchDarkly API access token once, and it’s stored securely. No env vars, token management, or coding against the LaunchDarkly API necessary.
Works with multiple MCP clients
Use the LaunchDarkly MCP server endpoint in Claude Desktop, Cursor, or directly inside Gumloop agents. Same server URL, works with any MCP client.
Chain LaunchDarkly with 100+ other integrations
Combine LaunchDarkly with Slack, Jira, GitHub, Google Sheets, and other MCP integrations in a single AI agent. An agent can pull a release detail from one tool, update a flag in LaunchDarkly, and write the result back somewhere else.
Enterprise-grade and scalable
Built for teams, with role-based permissions and dedicated support for Pro users. For details on Gumloop’s security practices, see trust.gumloop.com.
Pricing includes a free plan
You can test the LaunchDarkly MCP integration on Gumloop’s free tier before committing. Paid plans start at $37/month.
Frequently asked questions
Ship LaunchDarkly agents in minutes
Connect any AI agent to 100+ MCP servers, fully hosted, zero setup.