• EXECUTION
  • OBSERVABILITY
  • MONITORING
  • HUMAN

How to trigger and debug pipelines with your AI

Connect Codefresh to your AI client once. After that, you can run builds and inspect logs using natural language.

Ask AI about this page

Short answer

Can I run and debug Codefresh builds in my AI chat?

Yes. You can trigger new builds and pull specific execution logs to find errors. This lets you identify why a pipeline failed without leaving your chat interface.

The outcomes

What you can do with Codefresh

Your AI handles the heavy lifting of interacting with the Codefresh API.

EXECUTION

Start new builds

Tell your AI to kick off a specific build. It handles the request and confirms when the process starts.

OBSERVABILITY

Pull detailed logs

When a build fails, your AI fetches the execution details. You get the specific error logs needed to fix the issue.

MONITORING

List recent builds

Ask for a list of recent activity. Your AI pulls the build history so you can track status or find specific IDs.

HUMAN

Fix the code

The AI identifies the failure point in the logs. You then step in to apply the actual code fix in your IDE.

The workflow

What your AI does when a build request arrives.

The AI translates your chat commands into direct API calls to Codefresh.

  1. Identify the target

    The AI looks up your recent pipeline activity to find the right build or project you're talking about.

    list_codefresh_builds

  2. Launch the build

    Once you confirm the parameters, the AI sends the command to start the build process.

    trigger_codefresh_build

  3. Fetch failure data

    If the build hits an error, the AI pulls the specific execution logs to see exactly where it broke.

    get_build_execution_details

  4. Report findings

    The AI summarizes the log output, highlighting the specific error message or failed step.

    get_build_execution_details

Try it

Copy these to start.

Use these prompts to interact with Codefresh immediately.

Starting points

Replace the bracketed text with your actual project or build names.

Codefresh Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_PdvyiuNy98nJgJhF2rWbswmgtQCEOaod7r7AsNCv/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Codefresh capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "codefresh-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_PdvyiuNy98nJgJhF2rWbswmgtQCEOaod7r7AsNCv/mcp"
    }
  }
}
Copy into chat04

  • List my most recent Codefresh builds.

  • Trigger a new build for the 'web-app' project.

  • Get the execution details for build ID 12345.

  • Why did the last build for the 'api-service' fail?

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Codefresh once, then ask.

Just link your account via the Vinkius dashboard. Your credentials stay encrypted, and you'll be ready to chat with your pipelines instantly.

Connect Codefresh to your AI

FAQ

How the Codefresh MCP behaves

  • 01

    Can the AI delete my Codefresh pipelines?

    No. This MCP only has tools to list, trigger, and read build details. It cannot delete or modify your pipeline configurations.

  • 02

    Can I change environment variables through the chat?

    No. The current tools are limited to triggering builds and retrieving execution logs. You must use the Codefresh UI to change settings.

  • 03

    Does the AI see my source code?

    No. The AI only sees the build execution logs and metadata provided by the Codefresh API.

  • 04

    Can I use this with Cursor or Claude Desktop?

    Yes. Any MCP-compatible client can use this once you've connected it through Vinkius.

  • 05

    How do I find a specific build ID?

    You can ask your AI to list recent builds, and it will provide the IDs you need to pull logs or trigger new runs.

  • More questions about Codefresh? The Connector page answers them. See everything the Codefresh Connector can do

Connect Codefresh to Claude, Cursor, ChatGPT & more