How to Use the Harness MCP in Pydantic AI

Q: How does Pydantic AI prevent invalid parameters when triggering a Harness pipeline?

The framework validates the arguments for executepipeline against strict schemas before calling the MCP endpoint. This catches malformed inputs before they reach your Harness deployment engine.

Q: Can I use Pydantic AI to parse Harness audit logs?

Yes, the data from getauditlogs is parsed directly into type-safe Python objects within Pydantic AI. This allows you to write robust assertions on your Harness platform's configuration history.

Run type-safe Harness CI/CD operations with Pydantic AI runtime validation.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Harness MCP to Pydantic AI

Create your Vinkius account to connect Harness to Pydantic AI and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Harness with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Type-safe pipeline execution with Pydantic AI

Parameters for `execute_pipeline` must match your schema exactly, and Pydantic AI enforces this before calling the MCP server. If your Pydantic AI agent tries to pass an invalid string or missing parameter, the framework raises a validation error immediately. This strict validation prevents broken Harness API calls and partial deployments. When the Pydantic AI agent checks progress with `get_execution_status`, the returned payload is parsed into typed Python models. You get complete type safety across your entire Harness deployment workflow. This keeps your automated processes predictable.

Validate Harness environments and microservices

Structured lists of target environments are returned by `list_environments` under strict runtime validation through the MCP channel. Pydantic AI enforces strict data models on this output, ensuring your agent never acts on corrupted or incomplete Harness environment configurations. This keeps your deployment targeting reliable. The Pydantic AI agent pairs this with `list_services` to verify that the target Harness service exists in the selected project. Because every Harness response is validated, there is no risk of the agent hallucinating a microservice name. The Pydantic AI code either runs with perfect data or fails cleanly.

Secure and structured secret verification

Credentials remain secure since `list_secrets` only returns metadata for verification. Pydantic AI parses the Harness metadata into validated schemas, allowing your agent to confirm that required secrets are configured. This prevents Harness runtime failures caused by missing environment variables. To ensure complete security, the Pydantic AI agent uses `list_projects` to verify it is querying the correct Harness organization scope. Every Harness project ID is validated against your Pydantic models. This ensures your Pydantic AI scripts never execute against the wrong Harness environment.

Setup guide

Set up Harness MCP in Pydantic AI

Prerequisites

Python 3.10+ installed
pydantic-ai-slim[fastmcp] package
Active Vinkius subscription with a valid endpoint token

1

Install Pydantic AI with FastMCP
Run pip install "pydantic-ai-slim[fastmcp]". The FastMCP toolset replaces the deprecated MCPServerHTTP class with full protocol support.
2

Configure the FastMCPToolset
Pass a JSON-style config dict to FastMCPToolset with your Vinkius URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. Supports Streamable HTTP, SSE, and Stdio transports.
3

Create and run your agent
Pass the toolset to Agent(toolsets=[toolset]) and call agent.run(). Swap openai:gpt-4o for any supported model — Anthropic, Google, Mistral, or Groq.

agent.py

from pydantic_ai import Agent
from pydantic_ai.toolsets.fastmcp import FastMCPToolset

toolset = FastMCPToolset({
    "mcpServers": {
        "harness-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

agent = Agent(
    "openai:gpt-4o",
    toolsets=[toolset],
    system_prompt="You have access to Harness tools.",
)

result = await agent.run("List recent Harness transactions")
print(result.output)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Harness. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Harness now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Harness MCP in Pydantic AI

The framework validates the arguments for `execute_pipeline` against strict schemas before calling the MCP endpoint. This catches malformed inputs before they reach your Harness deployment engine.

Yes, the data from `get_audit_logs` is parsed directly into type-safe Python objects within Pydantic AI. This allows you to write robust assertions on your Harness platform's configuration history.

Yes, the server integration works with Pydantic AI's async runtime. You can poll `get_execution_status` concurrently without blocking your main Harness automation thread.

Pydantic AI will raise a validation error instead of passing dirty Harness data to your agent. This prevents silent failures and keeps your Harness CI/CD automation predictable.

The server never retrieves plaintext secrets, only metadata via `list_secrets`. All communication is encrypted, and Harness project structures from `list_projects` are processed locally within your secure Pydantic AI runtime.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript