Free in every series

Start with a free workflow: run one command, read a finished report.

Every series begins with a free L0 workflow. Unzip it, run the bundled example and read the report before you use your own data. No payment. In our store, enter your email address, accept the terms and we send you the download link. Or get it on Gumroad.

Choose your free workflow Compare all 25 workflows

1,402 automated tests passing in our run across the 25 packages, 25 of 25 first results written.

One per series

The five free L0 workflows

Each one is a complete workflow: runnable code, a bundled example, an HTML report, automated tests and a guidebook. Pick the problem in front of you.

S1 · AI-Native QA & SDET · L0 FREE
Failure Evidence Triage Workflow
Turn one noisy test failure into a clear next step, and see which parts are facts and which are guesses.
Best for: Anyone with a fresh failing test who wants to think clearly before changing code or escalating. You need: Node.js 22 and a terminal.
- A failure record in which every fact points to the log line it came from.
- Your own hypotheses listed apart from the evidence, with any conflict between sources kept in view.
- One next action for a person to approve, marked as not run until someone actually runs it.
Also inside (2)
- An HTML triage report from one command on the bundled sample, plus English and Polish guidebooks including a 15-page guide to auditing a failure bundle.
- A blank input template with a field guide, so you can run the same command on your own sanitized failure, and 25 automated tests.
Get it free Also on Gumroad
Choose the checkout you prefer; receipts and updates come from where you get it.
Next level: Test Failure Triage Workflow (L1 START, $19). Turn one failed test into a source-linked handoff: the evidence, the candidate explanations to test and one next check a developer can act on.
See the full S1 · AI-Native QA & SDET page
S2 · AI Evaluation & Reliability · L0 FREE
LLM Failure Pattern Scan
Stop trusting one good answer: repeat a small trial and see which of five failure patterns show up.
Best for: Engineers about to evaluate an LLM feature who are not sure where to start. You need: Node.js 22 and a terminal.
- A report (JSON, Markdown and standalone HTML) built from the sanitized trials you record.
- The exact variation between repeated outputs, stated as a measured figure.
- The five failure patterns named in the guide: intermittent acceptance failure, context-condition failure, unsupported material claim, output-contract failure and system-version regression.
Also inside (2)
- The checks you declared run, and the report proposes one follow-up check.
- An 18-page guide, English and Polish guidebooks, a blank input template and 23 automated tests.
Get it free Also on Gumroad
Choose the checkout you prefer; receipts and updates come from where you get it.
Next level: Initial LLM Test Checklist (L1 START, $19). Run five repeatable checks on a promising LLM demo and write down the variation and the risks.
See the full S2 · AI Evaluation & Reliability page
S3 · AI Security Testing · L0 FREE
Initial LLM Boundary Review Workflow
Find the first security questions worth testing around an LLM, with written authorization and a named owner.
Best for: Engineers and QA practitioners starting an AI security review who want the scope and authority written down first. You need: Node.js 22 and a terminal.
- A boundary map that names what is in scope and who is responsible for it.
- A list of authorized checks built from inert, sanitized test cases.
- A finding-or-gap record for anything you observed, and the owner of the next action.
Also inside (3)
- An HTML report from one command on the bundled synthetic example, with English and Polish guidebooks and a 14-page guide.
- A blank input template with a field guide for your own scope. Left unfilled, it stops with a clear TEMPLATE_NOT_FILLED message and writes no report.
- 29 automated tests, including positive, negative and recovery cases.
Get it free Also on Gumroad
Choose the checkout you prefer; receipts and updates come from where you get it.
Next level: Authorized Prompt Injection Review (L1 START, $19). Review prompt-injection evidence you collected, with authorization and evidence limits written into the report.
See the full S3 · AI Security Testing page
S4 · AI Engineering & Orchestration · L0 FREE
Engineering Context Map
Know what an AI agent may read and change before it touches your codebase: a context and responsibility map with one safe next action.
Best for: Developers who want a clear map of a codebase and its owners before an AI agent starts to change it. You need: Node.js 22 and a terminal.
- A context map that keeps trusted and untrusted sources apart.
- A responsibility map with one named reviewer for the decision.
- A review decision that lists what is still missing, and one safe next action.
Also inside (2)
- An HTML, Markdown and JSON report from one command on the bundled example, plus a blank input template with a field guide for your own codebase, and English and Polish guides.
- 38 automated tests with positive, negative and recovery cases.
Get it free Also on Gumroad
Choose the checkout you prefer; receipts and updates come from where you get it.
Next level: Tool Boundary Review (L1 START, $19). Decide which tools and permissions an AI agent gets before it can act, and who approves each side effect.
See the full S4 · AI Engineering & Orchestration page
S5 · AI for Work & Business · L0 FREE
Work Problem Framing
Turn a vague, recurring work problem into a brief with a result you can recognize.
Best for: Professionals and business owners who need to turn a fuzzy work problem into a clear, reviewable brief. You need: Node.js 22 and a terminal.
- A task brief that states the recurring job and what a useful result looks like.
- An input boundary that names the sources you may use.
- A review question for the person who decides, and one safe next action.
Also inside (2)
- An HTML, Markdown and JSON report from one command, plus a ten-minute path that needs no programming.
- 45 automated tests, and English and Polish guides.
Get it free Also on Gumroad
Choose the checkout you prefer; receipts and updates come from where you get it.
Next level: Research Brief Builder (L1 START, $19). Write a research brief from a question and the sources you may use, with facts, inference and unknowns kept apart.
See the full S5 · AI for Work & Business page

If the free level fits

Go as deep as the job needs

Each series has four more levels: L1 START $19, L2 UNDERSTAND $49, L3 APPLY $99 and L4 SCALE $149, paid once in our store. Every level includes the ones before it, so L4 SCALE is the complete series.

Paid levels

7-day money-back guarantee, no questions asked, as set out in the terms.
At least six updates in each 12-month period for five years, through the platform where you bought.
What you paid for any lower level of the same series counts toward any higher level (L1 START at $19 makes L4 SCALE cost $130).

Before you decide

Our store asks for your email address and sends you a secure login link to your files.
You need a terminal and Node.js 22, installed once.
Reports describe the evidence you supply and list what was not checked. No customer results are claimed: buyer usefulness has not been measured yet.

See the five levels and prices Compare all 25 workflows

Start with a free workflow: run one command, read a finished report.

The five free L0 workflows

Failure Evidence Triage Workflow

LLM Failure Pattern Scan

Initial LLM Boundary Review Workflow

Engineering Context Map

Work Problem Framing

Go as deep as the job needs

Paid levels

Before you decide

AI that helps with everyday work.