# Semaphore MCP for AI Agents AI Agent Connect > Semaphore MCP. Automate your CI/CD pipelines and manage builds directly through your AI agent. Stop jumping between tabs to check build statuses or hunt for logs. You can trigger promotions, manage deployment targets, and handle agent configurations using natural language. It brings your entire delivery lifecycle into your primary workspace. ## Overview - **Category:** developer-tools - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_4boMMcM8niVqPugdtaODagPmB0iDjPIFB9fElanu/ai-agent-connect - **Tags:** ci-cd, automation, pipelines, deployment, workflow-management ## Description This Semaphore MCP connects your Semaphore account to your AI agent so you can manage your delivery lifecycle through natural conversation. Imagine you are in the middle of a coding session and a build fails. Usually, you would have to switch windows, find the right project, dig through the dashboard, and scroll through hundreds of lines of logs to find the error. With this Connector, you just ask your agent what happened. It pulls the logs for you instantly. You can tell it to rerun a specific workflow or promote a build to production without ever leaving your editor. If you need to organize your storage, you can check artifact retention or list your deployment targets on the fly. It makes the heavy lifting of DevOps feel like a conversation. By connecting this through the Vinkius catalog, you get a direct line to your delivery pipeline. It is about staying in your flow state while still having total control over how your code gets shipped. You can manage self-hosted agent types, create new deployment targets, and validate your pipeline YAML definitions before you ever hit push. It turns a fragmented process into a single, cohesive experience. Instead of manually tracking every step of a release, you can ask for a summary of your deployment history or check the status of active agents. This means you spend less time looking at status lights and more time actually shipping features. Whether you are debugging a flaky test or preparing a major production rollout, your AI agent handles the tedious navigation while you focus on the high-level logic. ## Tools ### delete_deployment_target Delete a deployment target from your project. Use this to remove old or decommissioned targets. ### create_agent_type Create a new self-hosted agent type for your organization. This helps you define how your agents behave. ### get_workflow See the details of a specific workflow. Use this to understand how a specific process is mapped. ### list_agent_types See all self-hosted agent types in your account. This gives you a high level view of your infrastructure. ### rerun_workflow Start a rerun for a specific workflow. Use this to try a failed task again without manual setup. ### run_workflow Trigger a brand new Semaphore workflow. This lets you start a new process with a single command. ### update_agent_type Modify the settings for an existing agent type. Use this to change runner configurations on the fly. ### update_deployment_target Change the details of a deployment target. This allows you to update your environment definitions. ### create_deployment_target Set up a new target for your deployments. Use this to define where your code goes next. ### deactivate_deployment_target Turn off a specific deployment target. This is useful for maintenance or rotating environments. ### delete_agent_type Remove an existing self-hosted agent type. Clean up your configuration when types are no longer needed. ### disable_all_agent_types Turn off all agents of a specific type at once. This is a fast way to pause specific agent clusters. ### get_agent View details for a specific self-hosted agent. Use this to check the status of a single runner. ### get_agent_type See the configuration for a specific agent type. This shows you the settings for your runner groups. ### get_artifact_retention Check how long artifacts are kept for a project. This helps you manage your storage costs. ### get_deployment_history View the history of deployments for a specific target. This is perfect for auditing past releases. ### get_deployment_target_by_name Find a deployment target using its name. This makes it easier to identify targets in a large project. ### get_deployment_target Look up a specific deployment target by its ID. Use this for precise targeting in your scripts. ### get_job_logs Pull logs for a specific job to find errors. This saves you from scrolling through the dashboard. ### get_job Get the full description of a specific job. Use this to see the details of a single build task. ### get_pipeline View the details of a specific pipeline. This shows you the structure of your delivery flow. ### list_agents List all agents grouped by their type. Use this to see which runners are currently active. ### list_artifacts See all artifacts for a specific scope. This helps you track the files generated by your builds. ### list_deployment_targets See all deployment targets for a project. Use this to see all available environments for your app. ### list_pipelines List all pipelines for a given project. This helps you see every available delivery path. ### list_promotions View all promotions for a specific pipeline. Use this to track the progress of a release. ### list_workflows List all workflows for a project. This gives you a full overview of your automated tasks. ### rebuild_pipeline Fix failed blocks in a pipeline automatically. This helps you recover from partial failures quickly. ### stop_job Stop a job that is currently running. Use this to kill a hung process or a build that is taking too long. ### stop_pipeline Stop a running pipeline immediately. This is your emergency brake for any active delivery flow. ### stop_workflow Stop a running workflow. Use this to halt a specific process that is no longer needed. ### trigger_promotion Start a promotion to a new environment. This is the primary way to move code to production. ### trigger_task Run a specific task right now. Use this for on-demand actions like cache clearing or manual tests. ### validate_yaml Check your pipeline YAML for syntax errors. Use this to catch mistakes before you push your code. ### activate_deployment_target Turn on a specific deployment target. Use this to make a new environment available for use. ### configure_artifact_retention Set how long your build artifacts stay in storage. This helps you manage your storage costs and organization. ## Prompt Examples **Prompt:** ``` Show me the logs for job 889900 to see why the build failed. ``` **Response:** ``` Fetching logs for job 889900... **Job Summary:** - **Status:** Failed - **Error Type:** Timeout - **Step:** `npm install` **Log Snippet:** ```text [14:22:01] ERROR: Connection timed out. [14:22:05] Retrying... [14:22:10] ERROR: Registry unreachable. ``` It looks like a network issue with the registry. Would you like me to rerun the workflow? ``` **Prompt:** ``` List the last 5 workflows for project 'api-service'. ``` **Response:** ``` I've retrieved the latest workflows for project 'api-service'. | Workflow ID | Triggered On | Status | | :--- | :--- | :--- | | `wf-123` | `master` | Success | | `wf-122` | `develop` | Success | | `wf-121` | `feature-auth` | Failed | | `wf-120` | `master` | Success | | `wf-119` | `develop` | Success | Would you like to see the pipeline details for any of these? ``` **Prompt:** ``` Rerun the workflow with ID wf-554433. ``` **Response:** ``` I have triggered a rerun for workflow `wf-554433`. **New Workflow Created:** - **ID:** `wf-554434` - **Status:** Pending I'll monitor the progress for you and let you know if it completes successfully. ``` ## Capabilities ### Trigger promotions Move your builds to different environments with a simple command. ### Fetch build logs Get specific logs for any job to find errors quickly. ### Manage agent types Create and update self-hosted agent configurations. ### Control workflows List, run, or stop pipelines across your projects. ### Audit deployments View deployment history and targets in one place. ### Validate YAML Check your pipeline definitions for errors before they run. ## Use Cases ### Debugging a failed build A developer sees a failed build in their terminal and asks their agent to fetch logs for job 889900 to see exactly why it died. ### Promoting a release to production A release manager asks the agent to trigger_promotion for the latest pipeline to the production target to finish a deployment. ### Auditing past deployments A lead engineer asks the agent to show the deployment history of the staging target to see which version was last shipped. ### Managing storage costs A DevOps lead asks the agent to configure_artifact_retention to ensure that build logs and files are deleted after 30 days. ## Benefits - Stop context switching by checking build logs and status directly in your editor with get_job_logs. - Speed up releases by using trigger_promotion to move builds to production environments faster. - Fix broken builds quickly by using rebuild_pipeline to resolve failed blocks without manual intervention. - Keep your storage clean by managing artifact retention with configure_artifact_retention. - Maintain better infrastructure by managing agent types with create_agent_type and update_agent_type. - Catch configuration errors early by using validate_yaml to check your syntax before you push code. ## How It Works The bottom line is you get a command line for your CI/CD pipeline powered by natural language. 1. Subscribe to the Semaphore MCP on Vinkius. 2. Enter your Semaphore API Token and Organization Slug in your client settings. 3. Ask your agent to list workflows or check build logs. ## Frequently Asked Questions **Can Semaphore MCP help me debug failed builds?** Yes, it lets you pull specific build logs directly into your chat. You can identify errors without leaving your editor. **How do I use Semaphore MCP to see my deployment history?** You can simply ask your agent to show the history for a specific target. It will list all past deployments and their statuses. **Can I trigger a production release using Semaphore MCP?** Yes, you can use it to trigger promotions to your production environment using natural language commands. **Does Semaphore MCP work with my existing CI/CD pipelines?** It connects directly to your existing Semaphore account, giving your agent control over your current workflows. **How can I manage my build artifacts with Semaphore MCP?** You can list artifacts for specific scopes and configure how long they are kept in your storage. **Can I check my pipeline YAML for errors using Semaphore MCP?** Yes, it includes a tool to validate your pipeline YAML definitions before you commit them, helping you catch syntax errors early. **Can I check the logs of a specific job to debug a failure?** Yes! Use the `get_job_logs` tool with the Job ID. Your agent will retrieve the execution logs, allowing you to identify errors directly in the conversation. **How do I list all active pipelines for a specific project?** Simply ask the agent to run the `list_pipelines` action providing the Project ID. You can also filter by branch name or workflow ID. **Is it possible to trigger a promotion to production via chat?** Yes. If your pipeline has promotions configured, you can use the `trigger_promotion` tool by providing the Promotion ID to move your code to the next stage.