Where do you start from?
Quick show of hands
- Autocomplete and chat in the IDE
- An agent in the repository e.g. Claude Code with a
CLAUDE.md - An agent team e.g. subagents, skills, hooks
> /materialize ../my-project
Own the agent team that builds your way.
Bernhard Woditschka · github.com/woditschka/agentic-coding-reference
Where do you start from?
CLAUDE.mdHeard around the industry
What's in it for you
What you take from this talk. How such a team is designed and built, step by step.
Principal Software Engineer · Developing software is my passion.
Where the human sits
Prompting
my-service/ ├── src/ └── tests/
A rules file
my-service/ ├── CLAUDE.md build commands, conventions ├── src/ └── tests/
Then skills
my-service/ ├── CLAUDE.md ├── .claude/ │ └── skills/ how to test, review, ship ├── src/ └── tests/
Living specs
my-service/ ├── CLAUDE.md ├── .claude/ │ └── skills/ ├── docs/ │ ├── prd.md what to build │ ├── system-design.md how it is structured │ └── adr/ why it won ├── src/ └── tests/
A team of experts
my-service/ ├── CLAUDE.md ├── .claude/ │ ├── skills/ │ ├── agents/ requirements, design, review │ └── hooks/ what a script can enforce ├── docs/ │ ├── prd.md │ ├── system-design.md │ └── adr/ ├── src/ └── tests/
Automatic routing
my-service/ ├── CLAUDE.md ├── .claude/ │ ├── skills/ │ ├── agents/ │ └── hooks/ ├── docs/ │ ├── prd.md │ ├── system-design.md │ └── adr/ ├── .scratch/ │ └── handoff.jsonl a script picks the next hop ├── src/ └── tests/
The theorem
And once it is written, the humans benefit too.
The theorem
And once it is in code, the humans benefit too.
Eval task owners-page-param, harness v0.4.8, rep r2: 10 minutes end to end, $3.90 with grading, 29 ledger records. Walkthrough
What you get
Memory
Memory · the demo's fix
Long-term · docs/ · kept across features
docs/prd.md
A request for a page of owners numbered below one shows the first page instead of an error [REQ-OWN-005].
Done when: given a request for page zero or a negative page of the owner list, when it is answered, then the first page is listed as normal, not the error page.
docs/system-design.md
OwnerController · Server-rendered owner workflows: create, edit, search with paging, and detail · REQ-OWN-001, REQ-OWN-002, REQ-OWN-003, REQ-OWN-004, REQ-OWN-005
Short-term · .scratch/ · cleared after merge
.scratch/handoff.jsonl · line 3
{"type": "prd-entry", "author": "product-requirements-expert", "req_id": "REQ-OWN-005", "title": "Owner list requested below page one shows the first page", …}
Loops
Make it your own
/materialize installs it, upgrades it whole, keeps the open columnMake it your own · an example
Agree. The team decides: mocks at the service boundary are fine here.
Write it down. One paragraph in one brief; no agent file changes.
docs/testing-principles.md · shipped default
Mock only at system boundaries. HTTP clients, external APIs; these are the only acceptable mock points. Never mock internal code. Domain objects and services use real implementations. Hand-write mocks. No mock frameworks.
docs/testing-principles.md · your team's rewrite
Mock collaborators at the service boundary with Mockito. A service test stubs the repositories it calls. Never mock value objects. Construct them; they are one-line records. One mock library, the one already on the classpath.
Enforced. From the next slice on, the team writes and reviews to your rule.
What it delivers
| Eval task | Cost per pass | Wall | Bar |
|---|---|---|---|
| owners-page-param | $2.85 | 8 min | 3 of 3 |
| specialty-directory | $5.57 | 15 min | 3 of 3 |
| vets-specialty-filter | $5.23 | 15 min | 3 of 3 |
| visit-edit | $5.79 | 14 min | 3 of 3 |
Harness v0.4.8: cost per pass, median delivery wall, runs clearing the bar. evals/results/TREND.md
Deeper
Give agents what human teams already have.
One command. Your norms. The merge stays yours.
Slides and code: MIT License, © 2026 Bernhard Woditschka. Figures and recordings come from committed eval runs, provided as is, without warranty.
Built with reveal.js (MIT) and asciinema-player (Apache-2.0). Type: Inter and JetBrains Mono (SIL Open Font License 1.1). Page colors: Catppuccin Latte (MIT); terminal colors original. License texts ship in docs/deck/vendor/.
An independent project, not affiliated with or endorsed by Anthropic, GitHub, the OpenCode project, or the Spring project. Product names are trademarks of their owners, used only to identify the tools.
Views are the speaker's own. Legal notice and privacy.