← all play-ground
Live implementation

Agent Control Plane

This page is my own implementation of an agent control plane: a service I built and run that watches real deploys of another one of my projects, has an AI model read the actual code change behind each one, writes and runs browser tests against it, and publishes the result on a public dashboard. It's a working demonstration of AI-assisted software engineering, not a description of the concept.

Developers CI/CD Pipeline Agent Control Plane AI Agents Hosted LLM Providers · Local LLMs REST APIs · MCP Servers · Business Systems

The general pattern: this is how an agent control plane fits into a delivery pipeline, not a diagram of one specific deployment.

Live

Here's the real implementation.

Two cooperating services, both live right now: a small REST API that ships on every push, and a control plane that watches those deploys, reads the actual code change behind each one, writes and runs browser tests against it, and publishes the full result on a public dashboard, generated spec included. Everything below this point is real, not conceptual: every run on that dashboard reflects an actual push that happened, went through the pipeline described here, and reported in.

Real deploys, real tests

Every push becomes a tracked run with its own status and history, not a seeded demo dataset.

Diff-aware test generation

Specs are written against this deploy's actual diff and schema, not a static suite hand-written once.

Verified at the boundary

HMAC-signed webhooks, isolated databases, credentials that never touch version control.

Background: what an agent control plane is, and the problem it solves. The live implementation above is what this looks like built.

Concept

What is an Agent Control Plane?

A central platform for creating, deploying, managing and monitoring AI agents, so every team building on it shares one set of services (auth, orchestration, scheduling, governance, monitoring) instead of each reinventing its own. The full breakdown is in Capabilities below.

Concept

The Problem This Is Solving

Without a shared platform, every team ends up solving the same problems differently, and usually solving them worse the second and third time.

Without a control planeEvery team builds agents differently, with no shared pattern.
With a control planeOne platform, one pattern, reused across every team.
Without a control planePrompts get duplicated and drift out of sync across apps.
With a control planePrompts live in one place, versioned and shared.
Without a control planeAPI keys and secrets get copied into every project that needs them.
With a control planeSecrets are held once, centrally, and never touched by application code.
Without a control planeNo shared monitoring: agent failures surface as silent, unrelated bugs.
With a control planeEvery agent run is observable in one place.
Without a control planeNo auditing: nobody can answer what an agent did or why.
With a control planeEvery action is attributable and reviewable after the fact.
Without a control planeNo approval workflow: agents can act on production without a human in the loop.
With a control planeSensitive actions can require sign-off before they run.
Concept

Architecture

A general reference architecture for this category of platform, not a diagram of what's deployed below. See the "Live" architecture panel further down for the actual deployed system.

Developers GitHub CI/CD Agent Control Plane Agent Registry Agent Runtime Scheduler Workflow Engine AI Gateway Security Secrets Logging Memory API Gateway Agent Workers Hosted LLM Providers · Local LLMs · LLM Routers REST APIs · MCP Servers · Slack · GitHub · Jira · Databases
Concept

Agent Lifecycle

How an agent typically moves from an idea to something running in production, and staying observable once it's there.

01

Create

A developer defines the agent's task and boundaries.

02

Commit

The definition lands in version control like any other code.

03

CI/CD Validation

Automated checks confirm the agent behaves as specified.

04

Security Checks

Policy and permission review before anything reaches production.

05

Deploy

The agent becomes available for the control plane to run.

06

Run

The agent executes against real work.

07

Observe

Every run is logged and attributable.

08

Optimise

Cost, latency, and accuracy get tuned from real usage data.

09

Version

Changes ship as a new version, not an untracked overwrite.

Concept

Capabilities

The services a control plane typically provides, grouped by what they're actually for.

Orchestration & Runtime

Orchestration

Coordinates multi-step agent work across a workflow.

Scheduling

Runs agents on triggers, timers, or events.

Memory

Durable context carried across runs.

Governance & Security

Authentication

One shared identity layer for agents and callers.

Governance

Central policy over what agents can do.

Approval Workflow

Sensitive actions can require human sign-off.

Observability & Cost

Monitoring

Every agent run visible in one place.

Auditing

A reviewable record of what ran and why.

Cost Tracking

Spend attributed back to its source.

Integration

Model Routing

Selects which model handles a given request.

Integrations

A shared layer for connecting agents to external systems.

What building this actually involved

Two deployed services, one signed integration between them, and a pipeline that reads a real diff and turns it into a real, executed browser test. Underneath the AI steps, everything else, the webhook verification, the retry and repair policy, the database boundaries, the CI/CD pipeline, is deterministic code I wrote and can explain line by line.

Tech stack
TypeScript Fastify PostgreSQL Playwright Docker GitHub Actions