Build around evidence
Consume spread, local depth, imbalance, funding and eligible cross-market comparisons as named fields. Keep the source clocks, model revision and expiry alongside the observation instead of asking a language model to infer missing measurements.
Choose the workflow that fits
Agent developers can integrate the SDK; quantitative teams can inspect the calculation method; community operators can write explainable briefs from permitted outputs. Start with the public synthetic example before requesting project access.
Budget research separately from orders
A CatalystDesk project key authenticates the metered API. Credits pay for enabled data and model services; they do not grant trading approval. Execution rehearsal requires a configured Testnet account and independent approval. Descriptive context has no validated predictive performance.
An example of the questions to make explicit
Illustrative workflow · No market data or trading instruction
- Question
- What could change this market assessment?
- Evidence
- Source context and missing information are identified.
- Decision State
- Human review required
- Limitations
- No current market observation or execution instruction.
Check the service before integration
Check enabled capabilities in the authorized workspace before use. Public documentation does not activate a service. Simulation and testnet prototypes do not establish live trading availability.