All numbers on this page are generated from the public receipt, forward ledger, Playbook proof, and cycle index. The evaluator JSON is checked by verify:all.
Catalyst is an autonomous trading agent that knows when not to trade. Open the latest public receipt first: it shows Qwen's current thesis, the governor's approval or veto, and the checksum-backed record for that cycle.
Open latest receipt ->Bitget data feeds the market scan, Qwen generates the thesis, Agent Hub contributes context, and GetAgent proves the Playbook run. Remove those pieces and the loop is not Catalyst.
Open GetAgent proof ->Execution evidence is paper/backtest only. The claim is a reproducible perception -> decision -> risk -> receipt loop, not real-capital returns or guaranteed profit.
Open paper log ->Avoid chasing a move after an extreme 24h impulse.
Two win rates appear because they measure different proof surfaces, not conflicting totals.
Derived from committed approval receipts only. Balance and quantity are normalized paper-account values, not real capital.
| Time | Pair | Direction | Price | Quantity | Balance change | After | Receipt |
|---|---|---|---|---|---|---|---|
| Aug 19, 20:50 | SOLUSDT | Long | 74.5 | 0.232874 | +1.89 | 150.1 | ed620000...5da67a |
| Aug 19, 16:50 | BTCUSDT | Long | 63113.8 | 0.000284 | +1.58 | 148.2 | 49ccab55...89e34d |
| Aug 09, 20:54 | SOLUSDT | Long | 72.7 | 0.259515 | +1.17 | 146.7 | 7d8e33e0...9fab17 |
| Aug 09, 16:58 | SOLUSDT | Long | 72.8 | 0.249904 | +1.13 | 145.5 | de085db9...731f91 |
| Aug 06, 06:49 | ETHUSDT | Long | 1839.8 | 0.010957 | +0.79 | 144.4 | e72b7d77...7a5bc7 |
| Aug 06, 06:49 | BTCUSDT | Long | 62520.2 | 0.000309 | +0.70 | 143.6 | cf7c9d6d...f7dd09 |
| Jul 27, 17:58 | ETHUSDT | Long | 1862.9 | 0.008929 | +0.71 | 142.9 | 850def06...1c1c4e |
| Jul 27, 15:17 | ETHUSDT | Long | 1865.7 | 0.008878 | +0.60 | 142.2 | a494ed88...25ecda |
| Artifact | Purpose | Link |
|---|---|---|
| Evaluator proof JSON | One-file summary generated from all public proof surfaces | Open -> |
| Latest receipt | Current autonomous decision, checksum, thesis, council, and governor state | Open -> |
| Forward ledger | Autonomous cycle history and simulation-shadow discipline metrics | Open -> |
| Per-decision CSV | Downloadable rows for hand-checking discipline alpha | Open -> |
| Paper trading log JSON | Track 1 paper-trade rows with timestamp, pair, direction, price, quantity, and balance change | Open -> |
| Paper trading log CSV | Spreadsheet-friendly version of the Track 1 paper-trading log | Open -> |
| GetAgent Playbook proof | Completed GetAgent API run proof; no public rank claimed | Open -> |
Run the same gate used by the scheduled cycle workflow. The evaluator pack fails if it is hand-edited or stale relative to the source JSON.
npm cinpm testnpm run buildnpm run verify:allSimulation shadow, not real-capital PnL.
No profit guarantee and not financial advice.
Agent Hub fallback rows are labeled and do not count as live convergence.
The latest cycle can approve, veto, or stand down; the proof is the recorded decision path.
Verifier coverage: latest receipt checksum, decision-pair proof, discipline-alpha drift, calibration drift, council drift, paper-trading-log drift, evaluator-pack drift