CircleCI

Run the CircleCI MCP server as a Meshagent service and a chatbot in the cloud using the following steps:

1. Environment Variables

Required environment variables for the CircleCI MCP server:

• CIRCLECI_BASE_URL: (example: https://circleci.com)
• CIRCLECI_TOKEN: (replace with your CircleCI Personal API Token)

2. Installation & Authentication

• Install Meshagent with all enhancements:
  pip install "meshagent[all]"
  

• Sign up and authenticate with the Meshagent CLI.

3. Start the MCP Server as a Service in a Room

• Run the following command (replace env values as needed):
  meshagent service test --room=test --role=agent \
    --image=meshagent/mcp-circleci:latest \
    --env MESHAGENT_PORT=8001 \
    --env CIRCLECI_BASE_URL=https://circleci.com \
    --env CIRCLECI_TOKEN=your-circleci-token \
    --port="num=8001 path=/webhook liveness=/ type=meshagent.callable" \
    --name=mcp-circleci-service
  

  • This command starts a Meshagent room with the CircleCI MCP server available as a cloud service. Note: Rooms close if they go inactive.

4. Start a Chatbot in the Room (as a Tool User/Assistant)

• Join the room as a chatbot agent with access to the CircleCI toolkit:
  meshagent chatbot join --room=test --agent-name=sample --name=sample --toolkit=mcp-circleci
  

  • This starts a chatbot in the Meshagent room that can use the CircleCI toolkit. You can load multiple toolkits with the same agent if needed.
  • A link to the room will be shown in the command output—use it to interact with the agent and the CircleCI tools.

---

Tools Available

The following tools are available via the CircleCI MCP Server in Meshagent Rooms:

Tool	Description
config_helper	Analyze, validate, and fix CircleCI configuration files.
create_prompt_template	Generate a structured prompt template and context schema from requirements or existing prompts.
find_flaky_tests	Retrieve information about flaky tests in a CircleCI project; includes truncation warnings.
get_build_failure_logs	Get and analyze failure logs for CircleCI builds, with attention to possible log truncation.
get_job_test_results	Retrieve and filter CircleCI job test metadata and status; warns if output is truncated.
get_latest_pipeline_status	Get the status of the latest pipeline for a CircleCI project.
list_followed_projects	List all CircleCI projects that the user is following for convenient project selection.
recommend_prompt_template_tests	Generate recommended test cases for a given prompt template and context schema.
rerun_workflow	Rerun a CircleCI workflow from start or from the failed job.
run_pipeline	Trigger a new CircleCI pipeline and obtain a monitoring URL.

Tool Details

config_helper

• Purpose: Analyze and auto-fix CircleCI YAML config files.
• Input: configFile (string, raw CircleCI YAML).
• Output: Reports errors or confirms validity.

create_prompt_template

• Purpose: Generate structured prompt templates and input schemas from functional requirements or codebase prompts.
• Input: prompt, promptOrigin (“codebase” or “requirements”), model.
• Output: Template, schema, and prompt origin for use in prompt-based test/assist workflows.

find_flaky_tests

• Purpose: Identify flaky tests in projects or pipelines.
• Input: Provide either projectSlug, direct CircleCI URL, or workspace/git info.
• Behaviour: Checks for and handles truncated outputs.

get_build_failure_logs

• Purpose: Fetch and report CircleCI build failure logs.
• Input: Provide projectSlug & branch, URL, or workspace/git/branch.
• Note: Always handle truncation warnings.

get_job_test_results

• Purpose: Retrieve detailed test metadata and filter by pass/fail.
• Input: projectSlug & branch, job URL, or workspace/git/branch.
• Features: Truncation handling, result filtering (success, failure).

get_latest_pipeline_status

• Purpose: Get the status or progress of the latest CircleCI pipeline for a project/branch.
• Input: projectSlug & branch, project/pipeline/workflow/job URL, or workspace/git/branch.

list_followed_projects

• Purpose: List all CircleCI projects the current user follows; provides projectSlug for use by other tools.

recommend_prompt_template_tests

• Purpose: Generate recommended test cases for an LLM prompt template.
• Input: Prompt template, schema, prompt origin, model.

rerun_workflow

• Purpose: Rerun a CircleCI workflow by ID or direct workflow URL, from the failed job or from scratch.

run_pipeline

• Purpose: Start a new pipeline in CircleCI and get a link to monitor it.
• Input: projectSlug & branch, direct URL, or workspace/git/branch.

---

More Resources

• Meshagent homepage: https://www.meshagent.com/
• Getting Started with Meshagent Rooms: https://docs.meshagent.com/room_api/get_started
• Meshagent Room API Overview: https://docs.meshagent.com/room_api/overview
• Meshagent MCP Tool Docker Image: https://hub.docker.com/r/meshagent/mcp-circleci

---

With Meshagent and the CircleCI MCP Server, you can deploy powerful developer automation tools in collaborative cloud rooms—making AI-driven workflows available to your entire team, anywhere.