LaunchDarkly MCP Server for Claude
Manage LaunchDarkly feature flags, AI configs, environments, and audit logs from Claude — with the official LaunchDarkly Model Context Protocol server backed by the LaunchDarkly REST API.
Open the source and read safety notes before installing.
Safety notes
- Feature flag operations include create, update (patch), and delete — changes affect live targeting rules for all users in the environment.
- Audit log and code reference reads are non-destructive, but flag patch and delete operations are irreversible and will affect production traffic immediately.
- Scope the API key to the minimum required permissions (read-only for analysis workflows, write only for flag management workflows).
Privacy notes
- Feature flag definitions, targeting rules, user segments, AI config prompts, environment configurations, and audit log entries from your LaunchDarkly project are surfaced in Claude's context.
- Your API key is passed as a CLI argument — store it in Claude Code's MCP config rather than in shell history; rotate keys from the LaunchDarkly Authorization page.
Prerequisites
- A LaunchDarkly account and an API key from the LaunchDarkly Authorization page (Settings → Authorization → API Keys).
- Node.js (v18+) with `npx` available, and an MCP client such as Claude Code or Claude Desktop.
Schema details
- Install type
- cli
- Troubleshooting
- No
- Scope
- Source repo
- Estimated setup
- 5 minutes
- Difficulty
- beginner
- Website
- https://launchdarkly.com
Full copyable content
{
"mcpServers": {
"LaunchDarkly": {
"command": "npx",
"args": ["-y", "--package", "@launchdarkly/mcp-server", "--", "mcp", "start", "--api-key", "YOUR_API_KEY"]
}
}
}About this resource
Overview
The LaunchDarkly MCP Server is the official Model Context Protocol server from LaunchDarkly.
It gives Claude direct read/write access to your LaunchDarkly project — feature flags, AI configs,
environments, code references, and audit logs — through the LaunchDarkly REST API. It runs locally
via npx @launchdarkly/mcp-server over stdio, authenticated with a LaunchDarkly API key. Licensed
under MIT.
Key capabilities
- Feature flags — list, create, inspect, update targeting rules (patch), and delete feature flags.
- AI Configs — manage AI configuration definitions for model-backed features.
- Environments — list and inspect the environments in your LaunchDarkly project.
- Audit log — view the history of changes to your project's flags and settings.
- Code references — connect feature flag keys to their locations in source code.
Tools
| Resource | Operations |
|---|---|
| FeatureFlags | list, create, get, patch (update), delete |
| AiConfigs | list, create, get, delete |
| Environments | list, get |
| AuditLog | list |
| CodeReferences | list |
How it compares
| Server | Feature flags | AI configs | Environments | Audit log | Auth |
|---|---|---|---|---|---|
| LaunchDarkly MCP | Full CRUD | Yes | Yes | Yes | API key |
| Unleash MCP | Full CRUD | No | Yes | Limited | API key |
| Flagsmith MCP | Full CRUD | No | Yes | No | API key |
| Split.io MCP | Read + create | No | Yes | No | API key |
LaunchDarkly's MCP is the only feature flag server to also expose AI Config management, making it well-suited for teams using LaunchDarkly's AI model-config feature alongside standard flag-based rollouts.
Installation
Claude Code
claude mcp add launchdarkly -- npx -y --package @launchdarkly/mcp-server -- mcp start --api-key YOUR_API_KEY
Replace YOUR_API_KEY with an API key from Settings → Authorization → API Keys in your
LaunchDarkly account.
Claude Desktop
{
"mcpServers": {
"LaunchDarkly": {
"command": "npx",
"args": ["-y", "--package", "@launchdarkly/mcp-server", "--", "mcp", "start", "--api-key", "YOUR_API_KEY"]
}
}
}
Standalone binary
Download the pre-compiled binary from the GitHub Releases page:
curl -L -o mcp-server https://github.com/launchdarkly/mcp-server/releases/latest/download/mcp-server-bun-darwin-arm64
chmod +x mcp-server
Then reference the binary path in your MCP config instead of the npx command.
Requirements
- A LaunchDarkly account with an API key (Settings → Authorization).
- Node.js v18+ with
npx(for the npm install method). - An MCP client (Claude Code or Claude Desktop).
Security
- Use a dedicated API key with the minimum required scopes for your workflow.
- For read-only analysis, create a reader-role API key to prevent accidental writes.
- Flag patch and delete operations take effect immediately in the target environment.
Source Verification Notes
Verified on 2026-06-18:
- Official GitHub repository
launchdarkly/mcp-server(MIT) documents the@launchdarkly/mcp-servernpm package,npxinstall path,--api-keyflag, and available resource groups: FeatureFlags (list/create/get/patch/delete), AiConfigs (list/create/get/delete), AuditLog, CodeReferences, and Environments. - MCP server listing at
feluda.ai/mcp-servers/launchdarklyindependently confirms the feature flag management and API key authentication model. - Claude Code MCP documentation at
code.claude.com/docs/en/mcpdescribes the stdio connector setup pattern used above.
Source citations
Add this badge to your README
How it compares
LaunchDarkly MCP Server for Claude side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
| Field | LaunchDarkly MCP Server for Claude Manage LaunchDarkly feature flags, AI configs, environments, and audit logs from Claude — with the official LaunchDarkly Model Context Protocol server backed by the LaunchDarkly REST API. Open dossier | ConfigCat MCP Server for Claude Manage ConfigCat feature flags from Claude — create, update, and delete flags and targeting rules, manage environments, find and clean up stale flags, and audit change history — with the official ConfigCat MCP server and its 52 tools for the full ConfigCat Management API. Open dossier | Unleash MCP Server for Claude Create and manage Unleash feature flags from Claude — evaluate whether a change needs a flag, detect existing flags to prevent duplicates, set rollout strategies, toggle environments, and get cleanup guidance — with the official Unleash MCP server following Unleash best practices. Open dossier | GrowthBook MCP Server for Claude Manage A/B experiments and feature flags in GrowthBook from Claude — list and create experiments, manage feature flags, check experiment results, and interact with your GrowthBook workspace — with the official GrowthBook MCP server. Open dossier |
|---|---|---|---|---|
| Trust | ||||
| Install risk | Review first | Review first | Review first | Review first |
| Notes | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ |
| Category | mcp | mcp | mcp | mcp |
| Source | source-backed | source-backed | source-backed | source-backed |
| Author | LaunchDarkly | ConfigCat | Unleash | GrowthBook |
| Added | 2026-06-18 | 2026-06-18 | 2026-06-18 | 2026-06-18 |
| Platforms | Claude CodeCodexCursorClaude Desktop | Claude CodeClaude Desktop | Claude CodeClaude Desktop | Claude CodeClaude Desktop |
| Source repo | — | — | — | — |
| Safety notes | ✓Feature flag operations include create, update (patch), and delete — changes affect live targeting rules for all users in the environment. Audit log and code reference reads are non-destructive, but flag patch and delete operations are irreversible and will affect production traffic immediately. Scope the API key to the minimum required permissions (read-only for analysis workflows, write only for flag management workflows). | ✓Tools can create, update, and delete feature flags, targeting rules, environments, and segments — changes affect live feature flag configuration. Use `list-staleflags` before deleting flags to identify zombie flags and avoid breaking active SDKs. | ✓`toggle_flag_environment` enables or disables a feature flag in a production environment — confirm before toggling live flags. `create_flag` writes to your Unleash instance; capabilities are scoped by your PAT's permissions. Remote MCP (`--transport http`) is experimental and must be enabled on your Unleash instance. | ✓`create_experiment` and `create_feature` write to your GrowthBook workspace — capabilities are scoped by your API key/PAT permissions. Creating experiments or feature flags may immediately affect production if they are in a live environment — review before confirming. |
| Privacy notes | ✓Feature flag definitions, targeting rules, user segments, AI config prompts, environment configurations, and audit log entries from your LaunchDarkly project are surfaced in Claude's context. Your API key is passed as a CLI argument — store it in Claude Code's MCP config rather than in shell history; rotate keys from the LaunchDarkly Authorization page. | ✓Feature flag configurations, targeting rules, audience segments, SDK keys, and audit log entries from your ConfigCat account are surfaced in Claude's context. `CONFIGCAT_API_USER` and `CONFIGCAT_API_PASS` are Management API credentials — keep them in the MCP config env and never commit them to version control. | ✓Feature flag names, descriptions, targeting rules, rollout configurations, and project metadata from your Unleash instance are surfaced in Claude's context. Your `UNLEASH_PAT` is a secret — store it as an environment variable and do not commit it to version control. | ✓Experiment configurations, feature flag definitions, targeting rules, and experiment results from your GrowthBook workspace are surfaced in Claude's context. Your `GB_API_KEY` is a secret — store it as an environment variable, not in version-controlled config files. |
| Prerequisites |
|
|
|
|
| Install | | | | |
| Config | | | | |
| Citations | ||||
| Claim | Unclaimed | Unclaimed | Unclaimed | Unclaimed |
Signals
Loading live community signals…
A short, calm digest of reviewed Claude resources. Unsubscribe any time.