No RLXBT result yet
The strongest RLXBT use case is validating agent-owned imported market data. Mint a Manus token here, or let the first paid call return one, then retry after settlement.
RLXBT Daemon
Validate agent-owned market memory in RLXBT
RLXBT is meant for agents that need to spend, validate, and report on their own imported market history. Choose an agent-owned dataset, run a preset without raw rule JSON, then use scan, optimization, and report export as an auditable agent work product. The hosted source remains available only as a demo path.
Daemon health
ok
Version 0.1.1
Billing mode
Service-scoped Manus token
Tokenized agent spend, not a monthly subscription. The page retries paid execution with the same token_id.
Best-fit workflow
Agent-owned data first
Ready private-memory datasets are the primary agent lane. The hosted daemon dataset is useful for demos, but the strongest RLXBT value comes from testing an agent's imported market history.
Fastest path
Start with agent-owned data, then run, optimize, and export the result.
1
Step 1: Setup & Import
Import or choose an agent-owned dataset, then pick the preset the agent should validate.
2
Step 2: Settle Token
Mint a Manus backtest token, settle it in SOL, and reuse the same token across paid retries.
3
Step 3: Export Winner
Save the winning variant, inspect walk-forward stability, and export the report for another agent or orchestrator.
Launch panel
Start an RLXBT run
Recommended agent path
Start by onboarding a private dataset for an agent identity. Use the hosted RLXBT source only as a demo/reference lane.
This run will use the hosted daemon's mounted dataset for demo/reference only.
The highest-signal RLXBT workflow is dataset-first: import agent-owned candles, then use walk-forward validation and optimization on that history.
Walk-forward validation unlocks once you run optimization against a private dataset.
Windowed validation activates together with private-dataset optimization.
Agents should pass this as token_id on MCP calls or as X-Manus-Token on paid REST routes.
Scan runs all shipped presets against the current source and ranks them by fit so an agent can stop guessing which family to test next.
Private-memory export
Agent datasets for RLXBT
Use your ready private-memory datasets as RLXBT input files. Exports include full OHLCV-style columns; close-only datasets are normalized with flat open/high/low and zero volume so the RLXBT CSV loader can ingest them.
Sign in to see and export your private-memory datasets.
Run history
Saved RLXBT runs
Successful runs are stored automatically for signed-in users so you can reload them, compare presets, and avoid losing results on refresh.
Sign in to persist RLXBT runs across refresh and compare them later.