You know how to test but have no CI system, no evidence trail, and no answer when the EM asks: "Can we release?"
test executor
release decision lead
AI QA Engineer / Architect Path
AI can generate tests. Modern teams still need engineers who can stabilize CI, govern AI-generated testing, assess release risk, and make trusted ship decisions.
That is what this academy teaches — 8 operator playbooks, one cookbook repo, a capstone on your own product, and pass / warn / block gates your EM can sign off on.
Modern teams need engineers who can
8 operator playbooks · capstone on your repo · Pass / Warn / Block release gates
Why this academy exists
What most AI QA courses teach today
That gets you a demo and a chat thread — not a repeatable way to run QA when CI breaks every sprint.
Chat-only training still leaves teams unable to answer — with evidence:
You still use Claude and Cursor — but inside an playbook system: eight playbooks, one cookbook repo, artifacts leadership can audit.
Three starting points — one academy path. Same playbooks, same capstone.
You know how to test but have no CI system, no evidence trail, and no answer when the EM asks: "Can we release?"
You own Playwright and CI — but every release is still a debate and "green" doesn't mean confident.
You're using Copilot and Cursor daily — but there's still no owner, no fix order, and no repo evidence when Monday's CI breaks.
How you graduate
Eight playbooks · four release phases · one repo rhythm every sprint
Playwright · Cursor · Claude · LangGraph on playbooks
AI Test Automation Cookbook
Artifacts your EM can audit
Apply the cookbook to your actual codebase and deliver a release decision package your engineering leadership can review.
How the academy works
You do not need the product to learn. Modules 0–4 build skills in your forked cookbook repo. From Playbook 5 onward, optional TestNeo labs mirror the same playbook scenarios with team projects, impact analysis, and PR validation.
Fork the cookbook, wire IDE rules, run club:1–club:4. Build evidence in git — no product license required.
Module 0 · Playbooks 1–4Triage, gates, hygiene, command center — still anchored in repo artifacts and summary.json.
Playbooks 5–8Same release loop on the platform: generate from context, detect impact from diffs, run what matters, review before ship.
Capstone · optional platform trackNot for: One-off prompt experiments with no repo to keep — this is an playbook system you install on a real Playwright project.
Two tracks, one path
Both tracks share the same curriculum, capstone, and pass / warn / block framework. The distinction is how much AI QA Architecture work you take on.
Release playbooks
Each playbook module follows the same rhythm: short story → lab (npm run club:N) → artifact you keep. Playbooks 1–4 are cookbook-only; from 5–8 each card notes an optional TestNeo lab. Tap playbooks 1–8 in the hero preview.
Playbooks 1–4: cookbook repo only. Playbooks 5–8: same stories + optional TestNeo labs (MCP / extension) when you connect your trial.
Shortcut packs (repo commands)
club:greenFast path when smoke is the goalclub:redRed-suite triage when CI is on fireclub:tech-debtHygiene metrics + two-week debt planclub:listList all playbooks in the forked repoFork the cookbook, wire Cursor rules and AI rules of engagement, then connect basic MCP and LangGraph in your IDE — first bounded agent run on playbook context, output checked into repo reports. Optional: one testneo_validate_connection on TestNeo to preview the platform path.
You deliverForked repo · AI rules · Cursor workflow · MCP connected + first playbook-bound agent run
Tests fail: not found, timeout, stale selectors — you need green smoke evidence, not another rerun loop.
You deliverGreen smoke evidence
club:1 → 04 → 01 → 10 → 05
Optional flags: --with-visual (02), --with-semantic (03)
Stability Blueprint theme
UI is green but the API broke after a schema change — contract checks belong in the release story.
You deliverReadiness JSON / screenshot
club:2 → 06 → 07 → 12 → 20
Coverage gaps are obvious; the hard part is closing one theme without adding a flaky spec graveyard.
You deliverOne closed gap theme
club:3 → 11 → 13 → 14 → 29
Generation Blueprint theme
“Is it accessible / secure enough to ship?” — NFR work needs guardrails, not a last-minute scan.
You deliverNFR report + guardrails table
club:4 → 16 → 17 → 15 → 18 → 19
Monday 9am: hundreds of failures — one command to cluster themes and ship an RCA one-pager.
You deliverRCA one-pager
club:5 → 09 → 23 → 22 → 21 → 24 → 26 → 27
Ops Blueprint theme
“Can we ship Friday?” — convert noisy runs into a release decision doc with pass / warn / block.
You deliverRelease decision doc
Optional TestNeo lab: PR diff → impacted tests → testneo_validate_pr · PB 6 Preview in hero →The suite is slow, expensive, and messy — you need a two-week debt sprint leadership will fund.
You deliverTwo-week debt sprint
club:7 → 33 → 34 → 31 → 32 → 35 → 36
Governance Blueprint theme
Leadership wants one plan — not forty screenshots. Fuse upstream playbook reports into one riskScore and action plan.
You deliverExecutive slide with riskScore
club:8 → Recipe 37 orchestrator
Fuses upstream summaries from playbooks 5–7 (e.g. 28, 29, 33, 34, 36) — run after those clubs produce summary.json.
Run the playbook workflow on your codebase — same playbooks, real stakeholders. Choose how you prove graduate-level work:
club:N on your Playwright repo · summary.json evidence · ship / no-ship brief. No TestNeo license required.
Same release story using TestNeo: project setup, code impact, impacted test run or PR validation, dashboard findings, and a merge-ready summary for your EM.
You deliverRelease brief with pass / warn / block — repo paths, TestNeo workflow IDs, or both
In the repo: each recipe is a 9-section reference card (~20–30 min read). Playbook labs run npm run club:N in the order on each module card; Recipe 37 is the Playbook 8 orchestrator capstone (not a numbered cookbook chapter).
Recipes 01–10
Recipes 11–20
Recipes 21–28
Recipes 29–36
Playbook 8 capstone · npm run club:8 then run.py --auto-discover · fuses upstream summary.json into riskScore
| Module | Title | Hours | Deliverable |
|---|---|---|---|
| 0 | Repo, rules, MCP & LangGraph baseline | 2h | Fork repo · AI rules · MCP + LangGraph connected |
| 1–8 | Release playbooks | 2.5–3h each | Per-module artifact |
| Capstone | Release week | 4h | Ship/no-ship brief |
What you graduate with
Release decision tools, release playbooks, and leadership-ready evidence.
One concise brief that turns test results, risk, and ownership into a ship / hold / block decision.
Governance rules your team can reuse every sprint instead of debating release readiness from scratch.
Career outcomes
Different starting point. One destination — AI QA Architect.
Need this in your organization?
Learn → Implement → Scale
Graduate to TestNeo
Academy teaches how to operate in your repo. TestNeo is the engine for the same loop with projects, code-impact analysis, impacted runs, and shared history — without starting from scratch.
36 recipes + Recipe 37 · summary.json in git · pass / warn / block decisions · capstone on your codebase.
Same account as enroll · SDK, CLI, VS Code, MCP, and dashboard share projects, runs, and release signals.
The release loop
Each step has a cookbook lab and a matching product capability. You are not learning a different product; you are running the same verification intelligence with your team.
Bootcamp: Coverage maps, gap themes, reviewed specs in git
TestNeo: Projects, context, generate tests from code & design
Bootcamp: Diff-aware recipes, triage themes in reports
TestNeo: Code impact → impacted flows & tests
Bootcamp: club:N smoke, heal, targeted runs
TestNeo: Impacted-only runs, PR validation, IDE & CI
Bootcamp: summary.json, gates, EM briefs
TestNeo: Dashboard, failures, rerun, release signal
TestNeo is not another test runner. It is the verification layer between code changes and production releases.
Best for: Part 1–2 labs — NLP inside @playwright/test, optional publish to the team dashboard so repo and product share run history.
Recipes 01–14 · PB 1–4
Best for: In-editor test creation, impact detection from code changes, PR validation planning, and running impacted tests — same history as MCP and the web app.
Recipes 01–10 · change-aware validation
Best for: .testneo smoke and regression in terminal or CI — fast path when Playbook 1 “green smoke” should run before rewriting specs.
Recipes 03–06 · PB 1–2
Best for: Playbooks 5–8 in chat — testneo_validate_pr, triage, generate/run pipelines, release briefs (bounded tools, confirm=true for writes).
Recipes 21–36 · PB 5–8 · PR orchestration
Part 1–2 (01–14) — SDK, extension, or CLI for authoring and detection
Part 3 (21–28) — MCP + Web AI for triage, RCA, release signals
Part 4 (29–36) — MCP + dashboard for coverage, cost, governance briefs
Docs hub (SDK · CLI · MCP) · MCP prompt packs · Try TestNeo free
Pricing
Basic $59 — full path and capstone. Advance $119 — everything in Basic plus team-ops briefings and architect guides.
Self-paced · all 8 playbooks · ship the capstone
Everything in Basic + lead / architect track
Often called a RACI: for each playbook you name who runs it (R), who approves ship/no-ship (A), who you consult (C), and who gets informed (I) — so triage and release aren’t “everyone in Slack.”
FAQ
Contact: support@testneo.ai
Fork the cookbook, finish Module 0, run Playbook 1 on the sample app — one real artifact for your next standup. Then follow the full path to the capstone on your product.