howiclaude efficiency rubric — v1.0

"Efficiency" is a transparent 0–100 composite of five 0–20 sub-scores. Each is computed from what your config actually shows — or, where noted, from something you self-attested in the interview (those drive self-reported badges, marked as such on the page). A higher score is not "better engineer"; it's "more of the harness-optimization behaviors the community talks about." This rubric is versioned — every analyzed page is stamped rubricVersion: 1.0.

The five sub-scores (each 0–20)

• Context discipline — +8 if CLAUDE_CODE_AUTO_COMPACT_WINDOW is set at all; +4 more if it's tuned away from the default; +4 if you self-attest compacting before ~40% context; +4 if a Stop/PreCompact hook or a /compact-style command exists.
• Automation depth (weighted, never raw counts) — hooks: +3 for the first, +2 for the second, +1 each for #3–#6 (cap +9); slash commands with a non-trivial body: +1 each (cap +5); subagents: +1 each (cap +4); skills: +0.5 each (cap +2). Twenty unused commands is not a 20.
• CLAUDE.md specificity — Claude grades the file 0–20 against a published grading prompt: concrete rules with rationale, path-scoped rules, a rules/ directory → high; vague platitudes → low; empty/absent → 0.
• Parallelism — +6 if worktree.* is configured or worktrees are used; +6 if you self-attest 2 concurrent instances, +10 for 3, +14 for 4+; +6 for a /loop/overnight-agent pattern (cap 20).
• Permission hygiene — start at 20; −20 if --dangerously-skip-permissions is the norm — and permissions.defaultMode: "bypassPermissions" is that norm, flag or no flag; −6 if permissions.allow is effectively *-ish (overbroad); a documented sandbox config recovers up to +6 of a deduction. Auto mode is not a deduction: defaultMode: "auto" (Claude Code's default since 2026-08-14) runs a safety classifier over every action — it is the opposite of skipping permissions.

Composite = the sum of the five (0–100).

Score-neutral badges (added 2026-09)

These are read from the rig's published stack fields and change no sub-score — the 1.0 formulas above are unchanged. Badges resting on a declared setting or interview answer render as self-reported (◇); badges resting on a number measured from your session logs do not. ("Measured" means measured by the publish prompt on your machine — the site takes the number as published, the same way it takes your streak; it cannot re-count your logs.)

• Loop Engineer — one or more standing loops (/loop, /goal, a routine, a saved workflow, a Stop-hook loop) published. ◇
• Fleet Commander — agent teams or cross-session messaging on (◇), or a measured fan-out of 10+ subagents in one turn.
• Ultracoder — dynamic workflows / ultracode in use. ◇
• Fable-tier — default model on the Fable family (◇), or ≥ 50% of measured tokens on it.
• Auto-pilot — permissions.defaultMode: "auto". ◇
• Lean Context — global CLAUDE.md at or under 60 lines. ◇ (Score-neutral on purpose: the CLAUDE.md sub-score grades specificity, not size — both can be true.)
• Cache-maxxer — prompt-cache hit rate ≥ 70%, measured from session logs.

The percentile band

For v1 the band — Top 1% / Top 5% / Top 10% / Top 25% / Top 50% / Below median — is a fixed score→band lookup, calibrated on the catalogued seed setups, because there isn't a real population yet. It becomes a true percentile once enough real pages exist. We will never imply precision the data doesn't support.

Archetype

The Vanilla Purist if your config is within a small edit-distance of the stock defaults (a flex for minimalists). Otherwise The Dictator if you self-attest dictating more than typing. Otherwise it's named by your dominant naming sub-score — Context discipline → Context Surgeon; Automation depth → The Maximalist; CLAUDE.md specificity → The Architect (if you also self-attest plan-mode-first) else The Documenter; Parallelism → The Parallel Demon. (Good permission hygiene is expected, not an archetype.) If none of those four apply at all, you're The Minimalist Monk (or The Maximalist if your composite is high anyway). The archetype list isn't closed — re-running the prompt to change yours is fine; it nudges better practice.

How creator earnings work

Three rails, one ledger. Nothing on the manage page is an estimate: the cockpit shows real clicks on your links and real ledger entries.

• Amazon, your tag. Set your own Associates tag in the cockpit and every Amazon link on your rig carries it. Amazon pays you directly; we never see those orders and take nothing.
• Amazon, our tag. Rigs without a tag carry the site's tag. The site keeps that commission — Amazon's Participation Requirements forbid paying anyone for using its links.
• SaaS referrals. Tools with a referral program link through the site's enrolled account with your rig as the sub-id. When a referral pays, the ledger credits 80% of the commission to your rig and keeps 20%. If you have your own referral link for that program, yours is used instead and the site earns nothing.

Balances accrue in cents. Payouts run manually by the site owner once a balance crosses $10, to the PayPal or Stripe link (or the private email) you set as your payout method; each payout is recorded against your balance with its reference. Refunds reverse the matching credit. Payouts are made from confirmed commissions only — a commission stays pending until the program confirms it, because a refund before then would claw the money back.