# DevCycle MCP for AI Agents MCP

> DevCycle MCP lets your AI agent manage complex software configurations directly through natural conversation. You can list and search feature flags, monitor deployment environments (like Staging or Production), and track variable variations without writing a single line of configuration code. It gives immediate visibility into whether a feature is active for certain users or environments.

## Overview
- **Category:** developer-tools
- **Price:** Free
- **Tags:** feature-flags, experimentation, deployment-management, targeting-rules, environment-monitoring, software-testing

## Description

Managing product features often means juggling dozens of toggles and checking which environment has which version. DevCycle MCP lets your AI agent handle that complexity by connecting directly to the platform's core data. Instead of logging into multiple dashboards, you simply ask your AI client questions like, 'What is the status of the beta search feature in Production?' The system retrieves the full configuration details, showing targeting rules and variations instantly.

This capability doesn't require specialized scripting; it just requires conversation. When connected via Vinkius, your agent gains operational control, allowing you to audit environment settings or even update a flag status with simple text commands. It’s designed for engineers who need rapid, reliable answers about what code is running where.

## Tools

### get_environment_sdk_keys
Lists all necessary SDK keys for every environment within a project so you can verify connectivity.

### get_feature_flag_details
Retrieves the complete configuration, including targeting rules and variations, for one specific feature flag.

### get_project_details
Gets detailed information about a specific project within your DevCycle account.

### list_active_flags
Identifies and lists every feature flag that is currently enabled for use in the project.

### list_project_environments
Shows all available deployment environments, such as Staging or Production, for a given project.

### list_devcycle_projects
Lists every DevCycle project associated with your main account.

### list_feature_variables
Shows all variables defined within a specific project, ensuring technical consistency across the board.

### list_feature_flags
Retrieves a list of all feature flags associated with a particular project and their current operational statuses.

### search_feature_flags
Searches for feature flags within a project using keywords or partial names to narrow down results.

### update_feature_flag_status
Changes the operational status of a flag—marking it as active or archiving it entirely—directly through chat.

## Prompt Examples

**Prompt:** 
```
Show me the configuration for 'New-Search-Algorithm' flag.
```

**Response:** 
```
**Feature Flag Details: New-Search-Algorithm**

*   **Status:** Active
*   **Variations:** Control (False), Treatment (True)
*   **Targeting:** 50% of users in Production.
*   **Rules:** Requires user role 'premium'. Specific User IDs: [user_901, user_902].
```

**Prompt:** 
```
What are the SDK keys for our Staging environment?
```

**Response:** 
```
Here are the required keys for your Staging deployment:

*   📱 Mobile Key: `mob_stage_abc123`
*   💻 Client Key: `client_stage_xyz789`
*   ⚙️ Server Key: `server_stage_qwe456`

Do you need me to copy the client key for your team?
```

**Prompt:** 
```
List all feature flags and environments for 'Main-App'.
```

**Response:** 
```
**Project: Main-App**

*   **Environments Found:** [Production, Staging, Development]
*   **Active Flags:** 🟢 New-Dashboard-UI (Targeting 10%)
*   **Archived Flags:** 🟠 Legacy-Search (No users affected).

You can update the status of any flag with a simple command.
```

## Capabilities

### Get project overviews
Retrieves high-level details and configurations for any specific DevCycle project.

### List all active features
Identifies every feature flag currently marked as active across your accounts.

### Get detailed flag configuration
Provides the full setup, targeting rules, and status for a single chosen feature flag.

### View all project environments
Lists every deployment stage (e.g., Development, Staging) available for your projects.

### Retrieve environment SDK keys
Fetches the necessary Software Development Kit keys tailored to a specific environment and platform.

### Search feature flags by name or tag
Finds relevant feature flags within a project using descriptive keywords.

### Change flag status remotely
Updates the operational state of a feature flag, changing it between active and archived directly via chat.

## Use Cases

### Debugging unexpected flag behavior in Production
A Software Engineer asks their agent: 'Why isn't the beta dashboard visible for user ID 123?' The agent uses get_feature_flag_details to pinpoint that the flag is only targeting users in a different region, solving the bug immediately.

### Auditing environment readiness before launch
A DevOps engineer needs to confirm all keys are ready for the next deployment. They ask for 'all Production SDK keys' and get_environment_sdk_keys delivers a list, confirming system readiness.

### Disabling a problematic feature quickly
A Product Manager sees a new rollout causing errors. They instruct the agent to use update_feature_flag_status on the faulty flag. The status flips from 'Active' to 'Archived', instantly stopping the bad behavior.

### Checking project structure and variables
A new team member needs context on a legacy feature. They ask for all list_feature_variables, which provides a clean inventory of every variable defined in the current project scope.

## Benefits

- Immediate auditability: Use get_feature_flag_details to pull the exact targeting rules for any flag, saving time compared to digging through UI menus.
- Safe deployments: Monitor project environments using list_project_environments before pushing code. Know exactly if you're working in Staging or Production.
- Operational control: Change a feature flag's status (active/archived) instantly with update_feature_flag_status, allowing for rapid kill switches during an incident.
- Key retrieval efficiency: Get the required SDK keys for any environment using get_environment_sdk_keys; no more manual lookups in separate consoles.
- Comprehensive visibility: List all flags and project details via list_feature_flags and get_project_details, giving a full picture of your product's configuration state.

## How It Works

The bottom line is you get instant, conversational access to your entire software release configuration without manual dashboard navigation.

1. Connect your AI client to this MCP using your DevCycle Management API credentials.
2. Tell your agent exactly what you need—for example, 'What are the SDK keys for Production?' or 'List all active flags in Project X.'
3. The agent sends a request to DevCycle and returns structured data: flag statuses, environment details, or required security keys.

## Frequently Asked Questions

**How does DevCycle MCP help me check feature flags?**
DevCycle MCP lets you ask your agent specific questions about flags, like 'What is the status of Feature X?' It doesn't just give a yes/no answer; it gives the full configuration details, including which users or environments see it.

**Can DevCycle MCP help me manage my deployment keys?**
Yes. You can use this MCP to list all available project environments and retrieve the correct SDK keys for any of those stages (Staging, Production). This saves you from jumping between key management consoles.

**What if I need to turn a feature off immediately?**
You can use this MCP to update a flag's status directly via chat. Instead of needing an engineer to deploy emergency code, your agent flips the switch from 'Active' to 'Archived', stopping the feature instantly.

**Does DevCycle MCP handle multiple projects?**
Absolutely. The MCP can list all projects in your account and then drill down into specific ones. You get a centralized view of flags, environments, and variables across your entire software portfolio.

**Is this suitable for my CI/CD pipelines?**
While it doesn't run the pipeline itself, you can use DevCycle MCP to verify pre-deployment requirements. For example, confirming all necessary environment keys are available before a merge.