Workshops: Agentic Coding by Tech Stack
Agentic Coding with Go: Workshop for Idiomatic Go Services
Go's compiler, gofmt and race detector give coding agents unusually fast and honest feedback. Yet agents still tend to write Go as if it were Java: deep interface hierarchies, ignored errors and goroutines with no way to stop. This workshop shows Go teams how to steer Claude Code, OpenAI Codex, GitHub Copilot or Cursor towards simple, idiomatic and concurrency-safe services.
Who is this workshop for?
- Go developers building backend services, CLIs or infrastructure tooling
- Platform and cloud-native teams whose tooling is written in Go
- Developers coming from Java, C# or Python who want agents to help without importing old habits
- Tech leads defining Go style and review standards
- Teams running Go services in Kubernetes environments
What you will take away
- Encode your Go style in an agent configuration: package layout, error wrapping, logging and context handling.
- Make go vet, staticcheck or golangci-lint and the race detector mandatory checks in the agent loop.
- Have agents write table-driven tests, fuzz tests and benchmarks that reflect real edge cases.
- Review agent-generated concurrency code for leaks, deadlocks and missing cancellation.
- Recognise over-engineering in generated Go code and ask the agent for the simpler version.
Agenda
Module 1
Go and agents: a good match with caveats
- Why compile speed and gofmt make Go agent-friendly
- Typical non-idiomatic patterns in generated Go code
- Claude Code, OpenAI Codex, GitHub Copilot and Cursor with Go modules
Module 2
Idiomatic Go as context
- Error handling, wrapping and sentinel errors
- Small interfaces, defined where they are used
- Exercise: refactor an agent-generated service into idiomatic Go
Module 3
Concurrency with care
- Goroutines, channels, errgroup and context cancellation
- Running go test -race in every agent loop
- Exercise: find and fix a goroutine leak with the agent
Module 4
Tests and benchmarks
- Table-driven tests and test helpers
- Native fuzzing for parsers and input handling
- Benchmarks and pprof before accepting performance claims
Module 5
Tooling and review
- A golangci-lint configuration as the shared team standard
- Dependencies: standard library first, govulncheck for known vulnerabilities
- A review checklist for agent pull requests in Go
Simple beats clever
Go's culture values plain code: explicit error handling, few abstractions and readable control flow. Agents trained on many languages often add layers Go developers would never write, such as repository interfaces with a single implementation or generic helpers for one use case. Each layer looks reasonable in isolation; together they make a service hard to read.
In the workshop we turn your team's taste into explicit, example-based rules and practise review techniques for asking the agent to simplify.
Concurrency needs evidence
Agents can write concurrent code that passes every test on a laptop and fails under load. That is why, in this area, we ask the agent for evidence rather than assertions.
- Race detector on every test run that touches concurrent code
- Every goroutine has an owner and a way to stop
- Context is passed through, never stored in structs
- Benchmarks and profiles instead of performance guesses
Set-up and delivery
We work with your Go modules or a prepared sample service, remotely or on-site, in English or German. neooutput AI Code Engineers guide the exercises, and optional follow-up coaching supports your first weeks of everyday use.
Frequently asked questions
Is the workshop suitable for teams new to Go?
It works best for teams who already write Go. Newcomers benefit too, because the idiomatic patterns we teach the agent are exactly the ones they need to internalise themselves.
Do agents understand generics in Go?
Yes, they can write and use generics. The more common problem is overuse; we discuss when generics help and when a concrete type is clearer.
Which coding agent is best for Go?
Claude Code, OpenAI Codex, GitHub Copilot and Cursor all handle Go well. The outcome depends more on the feedback loop of tests, linters and race detection than on the tool itself.
Workshop
Get in touch
Want agents to write Go your reviewers do not have to rewrite?
Tell us roughly how many people will join and pick dates that work – we will get back to you within 24 hours with a proposal for content, format and schedule.
Prefer to talk first?
15 minutes, free and without obligation – with our AI Trainers and AI Code Engineers.
Book a call for free (opens in new tab) +49 761 519 751 50 (24/7)Request a training / workshop
Workshop: Agentic Coding with Go: Workshop for Idiomatic Go Services · 1 day
More workshops
- Workshops: Agentic Coding by Tech StackAgentic Coding with Rust: Advanced Workshop
- Workshops: Agentic Coding by Tech StackAgentic Coding with PHP, Laravel & Symfony: Workshop
- Workshops: Agentic Coding by Tech StackAgentic Coding in TypeScript & Node.js Monorepos: Workshop
- Workshops: Agentic Coding by Tech StackAgentic Coding for Flutter & Dart: Workshop for App Teams
- Workshops: Agentic Coding by Tech StackReact Native & Expo with Coding Agents: From Code to Store
- Workshops: Agentic Engineering PracticesCoding Agents in Large Codebases & Monorepos: Workshop
