Actively developed.

We studied how players actually work in MonkerSolver, PioSOLVER, HRC, ICMIZER, and GTO Wizard — then rebuilt our workflow around what they do, minus the friction. New in both apps: Solve from here (spawn a child solve from any node with reach ranges, pot, stacks, and rake carried over — no copy-paste range juggling), spot templates, batch solving over board lists with texture-grouped aggregation reports and outlier detection, hand-history paste that prefills the spot, runout tables with per-card EV grouped by card class, lock baselines that show the EV delta of an exploit against equilibrium, one-click bounty A/B for PKO spots, practice from any node with cross-session stats, in-app preflop abstraction building with progress, and double-click / deep-link opening of spots and solution packs.

A ground-up native interface for the Mac: home library with recents and solution packs, guided solve creation, the full node browser with pot odds and SPR on the table, practice mode with EV-graded feedback, node locking, ICM view, and CSV/PNG export. The same engine underneath, at full speed.

Every change now builds and passes the full test suite on macOS, Windows, and Linux, and produces Windows binaries automatically.

Game switcher for PLO, PLO5, Omaha hi-lo, and multiway NLHE; tournament fields (PKO bounties, large-field stacks, FGS) with bubble factors displayed in-app; straddles, antes, and the full bet-sizing grammar.

Monte-Carlo EV per action for the query hand or any hand in range, reported as EV ± standard error; optional EV column for the top hands at the node. Computed in the background, cached until more training runs.

Live combo counts, token chips, a palette of common PLO macro families with suit/weight modifiers, top-X%-by-equity quick sets, and named ranges persisted across sessions.

Copy a prism:// link to the current node; opening it loads the spot and walks the browser straight there. The same line syntax works from the command line.

Per-action sliders with automatic rebalancing; lock a node, resume solving, and the remaining seats converge on the max-exploit counter. Locks round-trip through the spot file.

Sortable all-runouts frequency table at chance nodes, ←/→ runout stepping, and a clickable line strip of action and card chips for jumping around the tree.

Single docked window with re-arrangeable panels; the browser and practice mode render a real table graphic bound to the current node. Solution packs open in a read-only viewer mode.

Progressive bounties with fraction and carry-over (mystery bounties as expected draw value), expected-blinds future-game adjustment, and automatic bubble factors for ICM spots.

Five-card Omaha for turn and river spots, and full 8-or-better hi-lo split evaluation.

"Solve until stability <" with a live sparkline; training pauses itself with a Converged readout instead of guessing at iteration counts.