jevbooks

← All projects

instinct

Vasiniks/instinct

Balatro agent: code computes odds, Jev judges moves, gates decide.

Automated README summary

Automatically summarized from the README on 26 Oct 2026; not a run by jevbooks. Run instructions and author-reported outcomes still need verification.

What it is

A TypeScript agent that plays Balatro through the BlindDeck mod's JSON-RPC API, combining card math and simulations with a small decision model, plus a live dashboard for watching and replaying runs.

How it uses Jev

Each tick, code extracts legal moves and utilities, then asks TypeSafe Jev one System One request: an action Choice over moves, a position Score, a phase risk Noul, and at shops and joker packs a build Choice. The model's probability is gated (confidence ≥ 0.55, none-fits ≥ 0.2) and blended with the code prior before acting.

Primitives:choicescorenoul

Technique worth stealing

Gate the model's pick on calibrated confidence and none-probability, then blend its probability with a code-computed prior before acting.

Try it

node --env-file=.env src/server.ts, or node tools/batch.ts --games 20 --decks RED,BLUE,PLASMA --stake WHITE

View on GitHub

Start using Jev: API key and your first request →

Model judgments from README excerpts

Project signals use repository metadata, an existing summary and the first 700 README characters, not the full project. “Not found” means not detected in that material; it does not establish absence. These are the original judgments with recorded models and dates, not source audits or run results.

Models & evaluation dates

Inclusion gate
jev-1.14.0
Project signals
jev-1.14.0
Category
jev-1.14.0
Pattern tags
jev-1.14.0

Evidence

These signals use repository metadata, an existing summary and at most 700 cleaned README characters. ≥ 0.60 means reported in that excerpt; ≤ 0.40 means not found there. The middle band is inconclusive. None is a source-code audit or a test result. The runnable-tool question excludes research experiments; the recommendation question asks for a learning shortlist, not a quality verdict.

  • Jev is core to the projectFound in material0.69
  • Shows a System One patternFound in material0.95
  • Describes handling uncertaintyFound in material0.94
  • Reports measured numbersunclear0.48
  • Is a runnable toolFound in material0.70
  • Top-10 learning shortlist signalnot found in the excerpt0.22
  • Is a replica of the modelnot found in the excerpt0.10
  • Solves a common problem (0–2)score on a 0–2 scale0.05 / 2
  • About JevFound in material0.98
Signals
Jev is core to the project0.69Shows a System One pattern0.95Describes handling uncertainty0.94Reports measured numbers0.48Is a runnable tool0.70Top-10 learning shortlist signal0.22Is a replica of the model0.1000.51Solves a common problem (0–2)0.05 / 2012

no-judgment band (0.40–0.60)

Design patterns
Traffic light0.98Funnel0.93Scorecard0.75Blind review0.54Menu0.50Flight recorder0.35Inspection0.2900.51

no-judgment band (0.40–0.60)

The Jev model and date for each judgment stage are listed above. Card written by DeepSeek V4.1 Flash from the README on 26 Oct 2026.