Documentation Index
Fetch the complete documentation index at: https://mintlify.com/deskiziarecords/QUIMERIA-HYPERION/llms.txt
Use this file to discover all available pages before exploring further.
SMKPipeline.step() is the core method of the QUIMERIA-HYPERION backend. It processes one OHLCV bar through all 18 detector modules, the plugin layer, and the AEGIS bridge, then returns a single Python dict. This dict is the data contract between every consumer: the WebSocket sends it, the frontend renders it, and the AEGIS bridge reads r['veto']['trade_allowed'] before placing any order.
Stability rules
- Adding a new key is safe. Consumers ignore keys they do not recognise.
- Removing or renaming an existing key breaks the frontend. The sensor panel, execution panel, and AMD state display all read specific key paths by name.
- All values are plain Python types — no numpy scalars, no numpy arrays. The
_sanitize()function enforces this before any dict leaves the pipeline.
The execution gate
True means all Ring 0 conditions are satisfied. False means at least one veto condition fired; r["veto"]["reasons"] lists which ones.
Sample result dict
Field reference
Top-level envelope
The current OHLCV bar.
0-based index of this bar in the loaded dataset.
Total bars in the current dataset.
veto — the execution gate
Ring 0 veto decision. This is the canonical execution gate.
sensors — the frontend sensor panel
List of sensor rows rendered in the frontend. Contains 14 SMK sensor rows followed by one row per loaded plugin. Order is stable within a session.
execution — AEGIS output
Output from the AEGIS bridge. Only populated on PROCEED bars when the bridge is available.
amd, fusion, mandra
AMD state machine output.
Lambda fusion engine output (Ring 0 OBNFE).
Mandra Gate output (information energy gate).