Settlements
Close the proof loop — council settlements and universal reality loop (open → ripe → settled).
Operating Contract
- May ingest anything
- May name anything
- May forecast when bounded
- May settle when reality resolves
- May learn only from what was settled
- Users may ask; users may not write ground truth
- Must admit unknown, subjective, ungrounded, stale, or unscoreable
KELLY never promises 100% accuracy on uncertain reality.
100% transparency when proof receipt, unknowns, settlement plan, and replay path are disclosed.
First-win settles feeding the queue
Weather + sports desks toward 100 settled · process is not truth · total settled —
Engine → scoreboard bridge
Waiting for first weather feed settle tick (or rebridge)
No bridge run yet · Neon optional
Feed ticks settle forecasts; bridge mirrors them into durable precommits for the public scoreboard. No prime memory write.
Live proof mapruntime_probe_degraded_last_known_learning_visible
Follow KELLY's current governed learning lane from source to observation, receipt, settlement, calibration, replay, and governed memory. The headline metrics come from the live runtime; archived bootstrap lineage is kept only for traceability.
Sources
4
Observations
25
Hypotheses
1
Closures
353
Visible cells
20
Replayable
7
Total Neon cells
3260
Visible replay receipts
7
Next candidates
2
Velocity
moving_forward_candidates_visible
Source: live_neon. Blocker: none. Public users cannot directly write ground truth; archived ledger backlog is monitored separately from current cell replay receipts.
macro.world_...flation
Loop state: reality_closed_memory_eligible
Reality has closed this stream; replay or Court linkage decides memory eligibility.
Archived bootstrap lineage
mcell_p58_bfaec131d529f551
The archived bootstrap record is historical lineage. KELLY's current learning state is the live memory stream above, not this old checkpoint.
Memory status
eligible
Current authority
live_metrics
Next proof action
use_live_metrics_for_current_learning_status
Last checked: 09:21:10 UTC. Public user writes: blocked.
Settlement follow-throughERA 8 product loop
After reality closes, KELLY keeps the loop visible: calibration updates, replay links, and a clear return path so users can see what changed — without writing ground truth from the public UI.
Settled
Outcome recorded with source and timestamp
Scored
Calibration lane updates independently
Return
Replay + proof links stay one click away
Settlement queue dashboardCalibration drive queue for reality closure, receipt settlement, scoring, and replay readiness.ripe: 0pending: 0Unavailable
Queue stage
waiting
Forward proof accumulation queue
settled
0
target
30
pending
0
ripe
0
Drive progress
Settlement progress 0% toward the current forward-proof target.
Population
Scoreable rows 0 / total 0.
Next nudge
Resolve ripe proof receipts as source outcomes arrive.
No drive gaps flagged - forward proof is accumulating from the live calibration drive.