R-multiple
Research record
Current run, verified records.
Evidence progress, R movement, gate decisions, receipts, and safety posture follow the website-safe Ops feed. Controls stay behind restricted review.
Feed check
closes
posture
Action path
Ops status
Run board
Run board ready
R detail building
Week detail queued
Evidence sample building
Boundary active
Canonical sample gate 0%
Actions Action stream ready
- Evidence publishPROOF
Published receipts when present
- Receipt checkRCPT
Public-safe receipt counts
- Gate reviewGATE
Aggregate gate decisions
- Safety checkSAFE
View-only safety posture
- Candidate queueCAND
Candidate counts only
Public actions appear when published; controls stay protected.
- Position
- Ready
- Event time
- Visibility
- Aggregate only
R curve R total building
- Trades
- Awaiting count
- Wins
- Awaiting wins
- Signals
- Signal total open
- Decisions
- Decision total open
Public numbers are aggregate paper evidence, R-multiple progress, publish counts, gate counts, and safety posture. Dollars, accounts, routing identifiers, system lane identifiers, and controls stay behind restricted review.
Bars -
Evidence model
Receipts -
Published receipts
Signals -
Published count
Decisions -
Decision count
24h checks -
Runtime posture
View seal
Current view fingerprint.
- View ID
- -
- Visible
- 20
- Filter
- All records
- Focus
- System / 2026-06-17
- Evidence
- Sample summary
- Action
- Action source
Seal ready.
Records
20 visible records
ETA added a website-native Living Algo surface that translates operating signals into public-safe proof: architecture integrity, lifecycle classes, capital gates, runtime posture, and explicit redaction boundaries.
1 / 20
No records match this view.
- System The Living Algo: public trust surface over the architecture map ETA added a website-native Living Algo surface that translates operating signals into public-safe proof: architecture integrity, lifecycle classes, capital gates, runtime posture, and explicit redaction boundaries.
- Evidence Canonical edge evidence: live capital now requires broker-backed proof The launch path was narrowed to one binding rule: an edge must prove itself on broker-backed paper-live fills, then clear canonical evidence gate with at least 200 broker closes, edge-quality score at least 1.30, positive realized outcome, receipt matching, and positive edge. The old prop screen is now review only.
- System Living Architecture OS: self-updating structure map and public view ETA gained a living architecture surface: a 3D operator structure map, a self-updating restricted structure feed, and a sanitized public architecture view. It maps modules, services, strategies, APIs, database surfaces, and event flows as view-only structure; it does not expose secrets or controls.
- Evidence Restricted edge-engine persistence: observe-only memory for research decisions ETA added an isolated memory lane for edge-engine decisions: scoped records tables, protected review views, operator research status, and ops cards behind the control gate. The writer is observe-only/default-off and keeps sensitive action fields off the public surface.
- Runtime Public ops posture and protected cockpit: real status without public controls ETA shipped public-safe ops status, a protected cockpit, public ops routing, edge-search status, signed proof, and review remediation. The public surface publishes reachability, freshness, safe fleet posture, and receipts; controls and account details stay restricted.
- Governance Autonomous research canon: event spine, corpus guard, and supercharge rules The June 12 canon consolidated the architecture rules for the next ETA: event bus first, one authoritative model per concept, statistical governance by default, legacy lanes marked for retirement, and edge-lab corpus guards so contaminated data cannot silently drive decisions.
- Engine fix Data hygiene gate: from noisy 70/70 failure to real findings A stale schema expectation made every history CSV look broken: loaders expected timestamp while the documented data column was time. The fix aliased time to timestamp, turning noisy 70/70 failure into 62 real findings and 8 genuine survivors, including known synthetic-row corruption and futures market negative April-2020 prints.
- Safety VPS and broker paper-live readiness: fail-closed broker truth The broker lane moved to the right authority: VPS-owned gateway, bracket audit, supervisor/broker reconcile, failover checks, and paper-live readiness. By June 16, readiness also required fresh realtime entitlement evidence; missing or stale market-data proof blocks paper-live entry.
- System Elevated Terminal: one visual system across public web, restricted surfaces, and ops The public site, restricted surfaces, and ops surfaces were rebuilt into one coherent terminal-grade interface. The work was not cosmetic only: it paired the visual pass with broker-cost truth fixes, security headers, proof routing, and production surface contracts.
- Safety Operator safety cockpit: step-up halt control, gate tracker, review replay, and adaptation monitor The operator cockpit gained the controls and diagnostics a serious trading OS needs: a step-up-gated emergency halt control, a 14-gate "why no entry?" grid, Court of Appeals replay UI, and Bayesian adaptation/drift readout. Most surfaces are view-only; restricted mutation requires explicit step-up.
- Method June 3 macro input feeds: ECB, ETF flow, OPEC, and weather Four public data adapters were added to the strategy-input layer: ECB meeting calendar, spot-digital-asset market ETF AUM flow proxy, OPEC headline classification, and NOAA population-weighted HDD/CDD demand. They are public-safe research inputs, not proof of trades or promises of edge.
- Method Re-derivation sprint: 66 systems became pre-registered edge hypotheses After the deep review found no proven positive edge across the 66-system family, ETA switched from belief-driven tuning to pre-registered research: 12 edge archetypes, indexed system mappings, three first-principles hypotheses, and per-ticker baselines for the active futures universe.
- Runtime Broker-only routing and restricted read models The May 24 wave pinned live and paper-live execution routing to the active broker, disabled dormant broker lanes unless reactivated in code and docs together, and stabilized the restricted surface around server-side read models. The project identity shifted from futures market seed system to multi-ticker futures Trading OS.
- Evidence Wave-25 prop-launch infrastructure: the honest NO-GO verdict Wave-25 built the prop-launch and paper-live infrastructure, then produced the more important result: an honest NO-GO for the 2026-05-18 funded-evaluation evaluation. The system could observe, route, and report; it still did not have enough broker-backed edge evidence to justify promotion.
- Governance Wave-19 Force Multiplier: five-role AI coordination with cost and fallback controls The Force Multiplier layer became a five-role coordination system: Codex as lead architect, DeepSeek as high-volume worker, Gemini as second opinion and long-context lane, Hermes as operator bridge, and Claude as a disabled legacy lane. It routes development and ops work; it never owns broker connections or order routers.
- System Wave-18: autonomous operation stack, Kaizen, quantum, Hermes, and DeepSeek personas Wave-18 moved ETA from a strategy collection toward an autonomous operating system: routine autonomous mode for proven subsystems, a Kaizen improvement loop, supercharged quantum optimization, Hermes two-way operator bridge, and DeepSeek Reasoner/Steward/Executor personas. The important boundary: this is autonomy infrastructure, not live-capital permission.
- Method Closing the lucky-OOS-split trap: the IS-positive gate digital-asset market opening-range system at one config had agg OOS Sharpe +3.57 with DSR pass 77.8% - looked like a clean PASS. Per-window inspection showed IS Sharpe negative in 7 of 9 windows. The strategy lost money on its training data; the OOS positive was a lucky date split, not validated edge. Added an aggregate-IS-positive check to the gate. digital-asset market correctly flipped to FAIL. Same hole caught a second strategy's claimed PASS the same day.
- Promoted Fleet baselines: 5 strategies cleared the strict walk-forward gate In one engineering iteration the production fleet went from 2 to 5 promoted strategies - futures market opening-range, futures market opening-range, digital-asset market opening-range system (5y re-baseline), digital-asset market opening-range system, and futures market daily DRB. Every promotion is IS+ AND OOS+ on real walk-forward data with explicit per-fold accounting.
- Rejected Grid trading on directional crypto: an honest non-promotion Grid trading is a real strategy with a real edge - in the right market. After targeted optimization across digital-asset market, digital-asset market, digital-asset market, every timeframe from 1h to D, every spacing from 0.1pct to 5pct, every config produced edge-quality score below 1.0. The bracket-grid math is upside-down on directional crypto. We document the failure rather than ship a forced promotion.
- Engine fix Engine fix: Sharpe calculation FP-noise + deterministic-R guards Three identical -1pct returns produced sd=1.3e-17 (FP rounding, not a real signal), giving Sharpe = -1.2e+16. The walk-forward aggregator averaged that into 9 digital-asset market OOS folds and reported a fleet-wide -1.35e+15. Root-caused, fixed, regression test added. Same iteration caught a second variant: 4 wins all hitting identical target-result wins produce sd/mean ratio ~1.15e-5 - also meaningless for Sharpe. Guard tightened to 1e-3 relative dispersion.
Public boundary
No fake activity. No sensitive exposure.
This ledger publishes methodology, results, and freshness. Identity fields, financial fields, credentials, routing artifacts, and control actions stay off page behind the control gate.