Skip to content

Latest commit

 

History

History

README.md

Tutorials

A hands-on, progressive learning track for the Iterative Session Methodology. The other docs define the framework — this one makes you do it, one thing at a time, against a real project, with a checkpoint after each step.

The track is eight tutorials, T1 through T8. Alongside them: a worked transcript of a real session against the sample project, a reusable tutorial template, and a bundled sample project to practice on. New here? Start at T1 and follow the curriculum in order.

Why bother?

You can read the 9 principles, 6 phases, and 12 gates and still not know how to run a session. These tutorials close that gap: you install the framework, run one real session end to end, and see where the guardrails catch you. The payoff — fewer wrecked sessions, and session N+1 reliably better than session N — is documented, not promised:

How the series works

  • One objective per tutorial — mirrors the methodology's "1 deliverable per session." Each tutorial does exactly one thing.
  • Progressive — every tutorial names its predecessor, so the chain T1 → T2 → T5 → T3 → T4 → T6 → T7 → T8 builds on itself.
  • You produce a real artifact — an installed framework, a saved session doc, a feature you shipped, a near-miss you caught. Not read-along.
  • Cite, don't restate — tutorials link into BOOTSTRAP.md, SESSION_RUNNER.md, and ITERATIVE_METHODOLOGY.md at the right beat. One source of truth; the tutorials never fork the principles, phases, or failure-mode list.
  • Dogfooded — each tutorial is itself authored as a methodology session.

Two ways to practice — pick one per pass

Choose a single track for a given run (do one, then the other if you like — not both at once):

  • Track A — your own repo. Apply each step to a project you actually care about. The sample run is your reference for "what a good run looks like."
  • Track B — the bundled sample. Use the sample project (a tiny todo CLI with a real test suite and a backlog) and follow the steps step-for-step in a throwaway sandbox.

One running example threads the series: T1 installs the framework onto the sample project, T2 runs the first session (building one todo-CLI feature), T5 uses that same session's near-misses as its cautionary cases, T3 runs Session 2 — scoring Session 1's handoff and building the next feature — to make the compounding loop felt, T4 switches lenses, auditing that same CLI and drafting a custom workstream for it instead of forcing the work into Development, and T6 carries the repository-wide hardening that audit recommended across a planning → execution → consolidation campaign that no single session could produce, T7 zooms out from that single project to the whole portfolio — running the dashboard across many repos and letting the risk matrix and compliance grid decide where the next session goes — and T8 keeps the framework itself current, running bin/status/bin/sync to absorb a new canonical release without losing the project's customizations.

Curriculum

Eight tutorials, listed in series order. Each does one thing and names the next.

# Tutorial You'll be able to…
T1 Setup & First Bootstrap Install the framework into a project: root files, CLAUDE.md protocol block, task tracking, dashboard
T2 Your First Session, End-to-End Run one full 6-phase pass to one deliverable — with Phase 0 Orient and the Present gate front and center
T5 Cautionary Use Read the gates, the 26 failure modes, "1 and done", the vertical-slice gates, and the Plan-Mode exit trap — and know when the methodology is too heavy
T3 The Compounding Loop Run the handoff + scoring loop and watch session N+1 improve
T4 Choosing & Adapting a Workstream Pick the right workstream; spin up a custom one when none fits
T6 Multi-Session Campaigns Carry one deliverable across many sessions — planning → execution → consolidation, with a checkpoint at each boundary
T7 Portfolio & Dashboard Ops Run across many projects; read the health dashboard
T8 Keeping Adopters Current Use bin/status / bin/sync to stay in step with the canonical repo without losing customizations

Writing a tutorial

Start from TUTORIAL_TEMPLATE.md — it encodes the shared anatomy (front-matter, "You do X → Expected result" steps with checkpoints, FM-by-number callouts, the "Why this matters" hook, and the successor pointer).