Pair task tokens with meter movement—only when the pair is identifiable.
Agent Studio sees token turns, quota percentages, run timing, model, and thinking level, and now emits cached quota state at run boundaries. The missing bridge is a calibration-grade fresh and settled raw pair tied to the same run and window identity.
Observation inventory
The plan reuses current Studio signals. It does not ask TokenEconomy to probe a CLI or parse provider output.
| Observation | Status | Fields used | Forecast role |
|---|---|---|---|
| Per-turn bus token events | Exists | model; input/output; cache read/write; reliable run correlation on the rich path; optional timing | Build model-specific observed attempt consumption. Legacy events without an unambiguous run join are ineligible. |
| Quota snapshots | Exists | CLI/plan; fetched time; normalized used %; reset; raw/source; error/suspicious | Observe the provider meter and screen parser quality. |
| Run lifecycle | Exists | start/finish; duration; terminal outcome | Bracket events and detect reset crossing, overlap, and censoring. |
| Run configuration | Exists | exact model; thinking level; task metadata | Define calibration and task-history cohorts. |
| Calibration-grade quota at run start/end | Partial · fresh/settled pair missing | cached boundary state + age exists; still needs fresh raw used/remaining semantics, resolution, settled end, and same run/window ID | Cached boundary events are audit evidence, not sufficient calibration pairs; completion stays with the parallel card. |
Source anchors: Studio's bus bridge ↗, token usage contract ↗, and quota snapshot contract ↗.
From raw events to one calibration observation
Bracket
Raw start snapshot with meter direction and resolution, all run token events, terminal outcome, and an end snapshot after a defined settle policy.
Match scope
CLI, opaque account/plan, window kind, reset identity, exact model, and thinking level.
Exclude confounding
No reset/expiry, unknown account use, overlap, missing tokens, parser suspicion, or saturated meter endpoint.
Keep evidence
Append-only raw record, normalized record, parser/calibration version, and explicit exclusion reasons.
Estimate the slope, explain the equivalent capacity
First convert used/remaining provider readings into one increasing usedPct, while preserving the raw direction and resolution. Estimate percentage points per canonical observed token from those endpoint and token observations.
deltaPctᵢ = betascope × tokensᵢ + errorᵢ
effectiveWindowTokensscope = 100 / betascope
Fit directly
Use a positive, interval-aware weighted robust slope through the origin. Per-pair 100 × tokens / deltaPct ratios become unstable near a rounded zero; an unidentified slope returns uncalibrated.
Name honestly
The inverted result is an effective token-equivalent window under observed conditions—not an official provider token budget.
For a shared provider meter, model-specific equivalents describe all-one-model workloads against the same window. They are not separate model quotas.
Eligibility is stricter than “two timestamps”
A usable pair requires
- Same CLI/account-plan/window kind and compatible reset identity; a first-use anchored window needs a trusted pre-use state and explicit new anchor.
- No reset crossing and no unmodelled rolling-window expiry.
- No unknown concurrent account usage.
- Complete, non-duplicated tokens and known model/effort.
- A known outcome and settled, trusted end; saturation must be explicitly bound-coded for an interval-aware estimator.
Calibration campaign
Combine controlled, serial spans in fresh/quiet windows with opportunistic production spans that pass the same gates. Vary model, thinking, and task size, and collect evidence across multiple distinct windows.
controlledsame-windowserialmulti-windowRounding, parser glitches, and confidence
Quarantine before robust fitting
- Impossible values or sign/direction errors
- Unexplained downward jumps or reset changes
- Parser-suspicious or stale snapshots
- Incomplete tokens, overlaps, or unknown background use
Never silently winsorize a known parser fault; retain the raw record and reason.
Rounded zero and saturation are censored
An integer meter can remain at 17% after small real use; a floor/ceiling can clip a large delta. V1 derives endpoint intervals from known rounding—or a one-resolution error bound only when the capture contract can defend it—and fits against feasible delta intervals. Otherwise return uncalibrated.
Bootstrap distinct windows
Pairs from the same window are correlated. Resample complete window instances, diagnose a free intercept for background use, and validate on later time periods.
Confidence has several inputs
Distinct windows, movement vs quantization, recency, residuals, evidence quality, hold-out error, and coverage. Sparse or stale means low confidence or uncalibrated; a detected regime change closes the old calibration instead of blending it with new data.