LEVR Internal · Field Guide · Concept Map + Cards

Trailhead

Tom's system for making AI-built software compound: a shared team memory, a disciplined build loop, and the tools that carry work from idea to merged PR. Ten cards below teach the whole system — click a card to flip it.

01 · The System

Six pieces, one loop

lore
Shared memory. Git-synced vaults Claude searches before deciding and writes back into.
craft
The build loop. Brainstorm → spec → gauntlet → plan → execute.
camp
Multi-repo workspaces. One session coordinates platform + mobile + bridge.
portage
The carry to GitHub. PRs per repo, CI to green, ordered merges.
ranger
Background cleanup. Refines small captured tasks and drains the ready queue.
outpost
The lookout. Read-only dashboard of your PRs, CI, workspaces, reviews.
brainstorm → ADR / specs → gauntlet → plan → camp new → execute → portage PRs → merged  ·  lore remembers everything along the way
02 · The Deck

Ten cards, whole system

Front is a question — answer it in your head before you flip. That retrieval effort is what makes it stick.

Click or press Enter to flip · printing this page lays every front and back flat.

03 · First Week

Study plan — three sittings

Sitting one · meet the system (~25 min)

  1. (5 min) Open http://localhost:7313 and name every pane out loud — PRs, needs-review, workspaces, staleness. Move on when you can say what Outpost will never do (push work at you).
  2. (10 min) Run lore search for a topic you know well (e.g. "facilitator platform") and read three hits. Then flip cards 1, 2, and 10. Move on when you can explain where a lesson learned today lives next week.
  3. (10 min) Flip cards 3, 4, 5. Sketch the craft pipeline from memory on paper — stages and what locks between them. Move on when your sketch has the gauntlet before the lock, not after.

Sitting two · hands on (~30 min)

  1. (10 min) In a fresh terminal: camp new warmup from the platform repo, look at what appeared in the workspace, then camp remove --force --name warmup. Move on when you can say why worktrees mean parallel batches don't collide.
  2. (15 min) Start a fresh Claude session in any group repo and run /craft:brainstorm with a small real want. Stop when it produces a draft spec — don't build. Move on when the session has printed the exact next command to run.
  3. (5 min) End with /lore:flush, then lore sync. Flip cards 6–9. Move on when you can name what portage does that camp doesn't.

Sitting three · first real slice (~45 min)

  1. (10 min) Gauntlet the spec from sitting two in a fresh session — resolve or accept every finding. Move on when the doc is locked.
  2. (25 min) Plan it, then camp new a workspace, start Claude inside it, and run /craft:execute. Watch the first slice land. Move on when Outpost shows your workspace and its state.
  3. (10 min) Let portage open the PR(s). Answer the Say-it-back questions below from memory. Done when you get 3 of 4 without looking.

Note: learning order ≠ dependency order — you meet Outpost first because it's the cheapest door in, even though the loop it observes comes first logically.

04 · Cheat Sheet

Commands you'll actually type

MomentCommandWhere
Start a new piece of work/craft:brainstormClaude session
Review an ADR/spec/plan/craft:gauntletfresh Claude session
Turn a locked spec into slices/craft:planfresh Claude session
Build it/craft:executeClaude session inside a camp workspace
Make a cross-repo workspacecamp new <slug>terminal, from any group repo
Tear a workspace downcamp remove --force --name <slug>terminal
Search team memory/lore:search or lore searchClaude / terminal
End of every session/lore:flush then lore syncClaude / terminal
Open PRs + drive CI green/portage:pull_requestClaude session in workspace
Drain small captured tasks/ranger:refine · /ranger:executeClaude session
Is the dashboard up?trailhead outpost status / startterminal

Using it from VS Code

The CLIs (camp, lore, portage, ranger, trailhead) work in any new VS Code integrated terminal — the shell profile loads them. The /craft:*-style commands are Claude Code plugin skills: they appear in any Claude session started after install. The one rule that changes in VS Code: a session can't leave the folder it started in, so for execute work open the camp workspace folder itself (File → Open Folder → the path camp pwd <slug> prints, under ~/.local/state/camp/levr/worktrees/) and start Claude there — or run camp new in the integrated terminal, which drops you in, and launch claude from that terminal.

05 · Say It Back

Close the loop

1 · The spec and the ADR disagree. Which do you trust, and why?

The ADR — specs are point-in-time build orders; ADRs are the continuously current record of decisions. A stale spec is history, not instruction.

2 · Why does each craft stage get a fresh session?

Small contexts stay in the smart zone: cheaper, sharper, and no bleed-over — your planning isn't polluted by brainstorm context. Each stage ends by printing the next command; you run it clean.

3 · What must exist before /craft:execute can touch three repos at once?

A camp workspace (camp new) built from the levr group — one worktree per member repo — with the Claude session started inside it.

4 · Teach it: explain Trailhead to a teammate in two sentences tonight.

One good shape: "It's a shared memory plus an assembly line — Claude checks what the team already knows, designs under adversarial review, builds in isolated multi-repo workspaces, and walks its own PRs to green. You spend your time on the thinking stages; the building runs itself."