Watching One Session Is Not Ranking Every Operator

The short version: sessionwatcher

sessionwatcher is a session watcher for Claude Code. It monitors individual coding sessions for token usage — what you burned in this session, what the breakdown was. It does the per-session monitoring layer well: you see what happened in each session as it happens. But a session is a single lap, not a season. Watching one session tells you what happened there; it does not tell you how good the operator is across all their work.

SigRank takes token telemetry across all of an operator's sessions and asks a different question: how efficiently does this operator compound? 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. sessionwatcher watches the lap; SigRank ranks the driver. Both matter. Only one tells you who is winning.

Feature comparison

Why watching sessions isn't ranking operators

sessionwatcher answers "what happened in this session?" That is per-session monitoring, not operator evaluation. Two operators can have identical session reports and wildly different efficiency. One reuses cached context efficiently and produces 30K output tokens; the other re-sends the same context every turn and produces 3K. Same session, ten-fold difference in signal. sessionwatcher sees the same report either way.

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. sessionwatcher gives you the session report; SigRank tells you whether the operator's cascade is compounding or burning.

From session watching to operator ranking

If you already run sessionwatcher, you have the session telemetry. SigRank reads the same logs and adds the operator-level scoring layer session monitors 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 sessionwatcher alternative?

They operate at different scopes. sessionwatcher monitors individual Claude Code sessions for token usage — it watches one session at a time. SigRank scores operators across all their sessions and ranks them globally. sessionwatcher is a per-session monitor; SigRank is an operator-level scoring and ranking system. Watching one session is not the same as ranking every operator. You can run both — sessionwatcher for session-level awareness, SigRank for operator-level evaluation.

Why is watching sessions not the same as ranking operators?

Watching a session tells you what happened in that one session — how many tokens were burned, what the breakdown was. Ranking an operator tells you how efficiently they compound across all their sessions, where they stand globally, and whether their cascade is compounding or burning. A session is a single lap; an operator is a driver across a season. sessionwatcher watches the lap; SigRank ranks the driver. Both matter. Only one tells you who is winning.

What does sessionwatcher not measure that SigRank does?

sessionwatcher reports per-session token usage. SigRank derives the cascade architecture across an operator's entire body of work: 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). sessionwatcher tells you what happened in one session; SigRank tells you whether the operator's cascade is compounding or burning across all of them.

Can I use both sessionwatcher and SigRank?

Yes. sessionwatcher is a session-level monitor; SigRank is an operator-level scoring and ranking tool. Run sessionwatcher for per-session token usage awareness while you work. Run `sigrank submit` to get your Yield score, class tier, and leaderboard rank. They read from the same local session logs and do not conflict. sessionwatcher is for watching sessions; SigRank is for ranking operators.

Does sessionwatcher work with tools other than Claude Code?

No. sessionwatcher is a Claude Code session monitor — it reads Claude Code session logs specifically. SigRank is platform-neutral: it works across Claude Code, Cursor, GitHub Copilot, ChatGPT, Gemini, and 15+ other platforms. If you only use Claude Code, sessionwatcher is fine for session-level monitoring. If you use multiple tools and want a comparable operator-level score across all of them, SigRank is the answer.