Streaks Measure Consistency. Yield Measures Skill.

The short version: clawdboard

clawdboard aggregates token usage, layers on streak gamification and "cooking" project labels, and ranks operators by a blend of cost, tokens, streaks, and active days. It does the engagement layer well: clean UI, habit-forming streaks, a sense of progress. But it ranks by the wrong axis. Streaks measure attendance, not ability. An operator who shows up daily and re-sends the same context will have a perfect streak and a low signal-to-noise ratio. Higher streak, lower skill, higher rank.

SigRank takes the same four token pillars and asks a different question: is the cascade compounding or burning? The headline metric, Υ Yield = cache_read × output / input², rewards the operator who reuses cached context efficiently and penalizes the one who burns fresh input without leverage. clawdboard counts the days; SigRank grades the driving. Both matter. Only one tells you who is skilled.

Feature comparison

Why streaks and active days aren't enough

clawdboard answers "how consistent am I?" That is a habit question, not a skill question. Two operators can have the same 30-day streak and get wildly different outcomes. One reuses cached context efficiently and produces 30K output tokens per session; the other re-sends the same context every turn and produces 3K. Same streak, ten-fold difference in signal. On a streak leaderboard, they tie. On a Yield leaderboard, the gap is obvious.

SigRank's headline metric, Υ = cache_read × output / input², measures exactly that gap. It rewards the operator who compounds cached context into output and penalizes the one who burns fresh input without leverage. clawdboard counts the days you showed up; SigRank tells you whether the cascade you ran was compounding or burning.

From streaks to Yield

If you already run clawdboard, you have the logs. SigRank reads the same telemetry and adds the efficiency layer streak rankings never had:

npm install -g sigrank
sigrank enroll      # create your operator identity
sigrank submit      # reads logs, scores, signs, publishes

Prefer to inspect before you submit? Run sigrank me --dry-run to see your scored payload locally, or paste your token counts into the /score calculator to compute your Υ Yield, class tier, and compression ratio instantly, no account, no submission, just the numbers.

Frequently asked questions

Is SigRank a clawdboard alternative?

They overlap on data but diverge on the ranking axis. clawdboard aggregates token usage, adds streak gamification and "cooking" project labels, and ranks by a blend of cost, tokens, streaks, and active days. SigRank takes the same token telemetry and ranks by cascade efficiency: Υ = cache_read × output / input². If you want a habit tracker for AI coding, clawdboard is fun. If you want to know who is actually skilled, SigRank is the answer. You can run both, they read the same logs.

What does clawdboard not measure that SigRank does?

clawdboard reports cost, token totals, streaks, and active days, then blends them into a composite rank. SigRank derives the cascade architecture from the same four pillars: Υ Yield (is signal compounding or burning?), compression ratio (output per input), SNR (signal density), Leverage (how much cached context amplifies your input), and Velocity (tokens per unit time). clawdboard rewards showing up; SigRank rewards driving well. Streaks measure consistency; Yield measures skill.

Can I use both clawdboard and SigRank?

Yes, and they serve different purposes. clawdboard gives you the habit layer: streaks, active days, project labels, a clean UI that keeps you coming back. SigRank gives you the efficiency layer that streak counts cannot. Run `sigrank submit` to publish your cascade score to the SigRank leaderboard, and keep clawdboard for the motivation loop. The same local logs feed both.

Which is better for measuring AI coding skill?

SigRank. Streaks and active days measure attendance, not ability. An operator who shows up every day and re-sends the same context will have a perfect streak and a low Yield. An operator who shows up twice a week but compounds cached context into high-yield output will have a modest streak and a high Yield. If you want to find the best drivers, look at lap times, not how many days they showed up to the track.

clawdboard has "cooking" project labels. Does SigRank track projects?

SigRank tracks build archetypes, not project nicknames. The 10 build archetypes (e.g. Greenfield, Refactor, Debug Cascade, Migration) classify the kind of work the operator is doing, derived from the cascade signature rather than a self-applied label. This makes archetypes comparable across operators: a Debug Cascade run on one machine means the same thing as a Debug Cascade run on another. "Cooking" labels are fun but not comparable.