Starter
A first route with simple checks.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
NAVIGATE BY INTENT
From intent to evidence.
claude.guides by MATZPEN is a practical library for Claude Code and AI agents. Pick one real problem, take one small step, and leave a checkable result behind.
FIELD NOTES / FOR PEOPLE WHO BUILD & RUN
From intent to evidence.One checkable step.
Practical guides for Claude Code and AI agents. Each one gives you a small action, a clear success check, and a safe way to stop when the result misses the mark.
Choose where to start01 / THE LIBRARY
What problem is in front of you?
Choose a guide by the real problem at hand. Read the context, make a small change, and test it before widening the scope.
Builder
A workflow ready for a team.
Advanced
Boundaries, measurement and rollback.
04 / IN THE WORKSHOP
Paid guides
Not priced the way you might expect - your success should not have to fund our vacation. When this shelf opens, pricing will be clear, fair and tied to real value.
Agent Systems
A deep field guide to agent systems that teams can actually operate.
Context Architecture
How to design memory, handoffs and context without losing the plot.
Shipping Agents
From prototype to production with gates, ownership and rollback.
05 / OPEN LOOP
Requests & feedback
CHANNEL OPENING SOON
Missing a guide? Something unclear? Found an accessibility barrier? This is the loop we want to open for you.
Request a guide
Topics, workflows or patterns you would like to see in the library.
Guide feedback
What worked, what was missing and what should be clearer next time.
Accessibility report
A keyboard, screen reader, contrast or mobile barrier is a real bug, not a footnote.
Transparency before a button: There is currently no active email address or contact form, so this section does not collect or send information.
03 / SIGNALS
posthog / anonymous / inspectable
Real usage, without dashboard theatre.
Anonymous measurement only: no names, email addresses, search content, or personal identifiers. Data is stored in PostHog—not localStorage.
Events0
Opens0
Categories0
Awaiting the first signal
No usage data to show yet. Events are sent to PostHog anonymously.
DATA CONTRACT / ANONYMOUS ONLY
Anonymous measurement only: no names, email addresses, search content, or personal identifiers. Data is stored in PostHog—not localStorage.
02 / METHOD
Guides designed to leave the screen and enter the work.
Every guide starts with a problem you can name and ends with a next step you can test. Less prompt theatre; more work you can repeat.
01Define the context
02Run one small step
03Verify before scaling
04Record what you learned