decidekit 0.0.1 → 0.1.1

package/CHANGELOG.md

@@ -1,5 +1,26 @@
 # Changelog
 
+## 0.1.1 (2026-11-02)
+
+### Models
+
+- The default model id is `debuffjp/decide-small` (Decide Small). Before 0.1.1 it was `Xenova/bge-small-en-v1.5`. The
+  weights are identical, so decisions don't change. For other languages the id is `debuffjp/decide-multilingual`
+  (Decide Multilingual), identical to `Xenova/multilingual-e5-small`.
+- The old ids and heads recording them keep working. They resolve to the new ids, so `d.model` returns the new id. A
+  head trained on 0.1.1 records the new id, so a 0.1.0 decider needs that id passed as `model` to use it.
+- Downloads of both models are pinned to a revision.
+- New option `modelHost` serves model files from `<modelHost>/<model>/` instead of Hugging Face. It applies to the
+  whole page or process.
+
+### Docs
+
+- Label names count as 1 example in every doc, and the advice to turn them off with bge-small is gone: it applied to the
+  old weight of 3.
+- Calibration, out-of-scope and zero-shot numbers are now bge-small's or labelled with the model they were measured on.
+- `decide()` documents that a described label scores by the better of its name and description.
+- The API reference covers `decidekit/typesafe`.
+
 ## 0.1.0 (2026-11-02)
 
 The first release. Numbers are on public test splits unless noted; the scripts that produced them are in `eval/`.
@@ -47,7 +68,7 @@ The first release. Numbers are on public test splits unless noted; the scripts t
   171 s on the main thread; the longest stall in a worker was 18 ms).
 - `decidekit/react`: `useDecider()` and `useDecision()`.
 - `decidekit/switch`: a decider for the switch agent harness.
-- A CLI: `decidejs train` and `decidejs decide`.
+- A CLI: `decidekit train` and `decidekit decide`.
 
 ### Known limits
 

package/README.md

@@ -1,12 +1,12 @@
-# decidejs
+# decidekit
 
-[![npm](https://img.shields.io/npm/v/decidekit?color=cffb2d&labelColor=11140d)](https://www.npmjs.com/package/decidekit) [![ci](https://github.com/debuffjp/decidejs/actions/workflows/ci.yml/badge.svg)](https://github.com/debuffjp/decidejs/actions/workflows/ci.yml) [![MIT](https://img.shields.io/badge/licence-MIT-cffb2d?labelColor=11140d)](LICENSE) [![Discord](https://img.shields.io/badge/discord-debuff-a78bfa?labelColor=11140d)](https://discord.gg/jshbDHSpCY)
+[![npm](https://img.shields.io/npm/v/decidekit?color=cffb2d&labelColor=11140d)](https://www.npmjs.com/package/decidekit) [![ci](https://github.com/debuffjp/decidekit/actions/workflows/ci.yml/badge.svg)](https://github.com/debuffjp/decidekit/actions/workflows/ci.yml) [![MIT](https://img.shields.io/badge/licence-MIT-cffb2d?labelColor=11140d)](LICENSE) [![Discord](https://img.shields.io/badge/discord-debuff-a78bfa?labelColor=11140d)](https://discord.gg/jshbDHSpCY)
 
-**[Try it in your browser](https://debuff.jp/decidejs/playground/)** · [Watch it learn from an LLM](https://debuff.jp/decidejs/lanes/) · [Launch post](https://debuff.jp/blog/introducing-decidejs) · [Research: when the LLM is wrong](https://debuff.jp/research/learning-from-a-noisy-llm)
+**[Try it in your browser](https://debuff.jp/decide/playground/)** · [Watch it learn from an LLM](https://debuff.jp/decide/lanes/) · [Launch post](https://debuff.jp/blog/introducing-decide) · [Research: when the LLM is wrong](https://debuff.jp/research/learning-from-a-noisy-llm)
 
 A decision model for the browser and Node: it picks one of your labels for a piece of text, with confidence you can trust, and learns from an LLM when it isn't sure. Your labels are the type. No server, no API key.
 
-A System One model in the browser: like TypeSafe's Jev, but on the user's device, so each decision is free, private and a few milliseconds. Jev leaves unsure cases to your code; decidejs can ask an LLM about them and learn from the answers.
+A System One model in the browser: like TypeSafe's Jev, but on the user's device, so each decision is free, private and a few milliseconds. Jev leaves unsure cases to your code; decidekit can ask an LLM about them and learn from the answers. decidekit is an independent project, not affiliated with TypeSafe.
 
 ```ts
 import { createDecider } from "decidekit";
@@ -29,8 +29,8 @@ await d.decide(text, ["question", "bug_report", "praise"]);
 await d.decide(text, { refund: "wants their money back", bug: "something is broken" });
 ```
 
-Zero-shot is for trying things out: 61% on BANKING77 and 70% on CLINC150 (dev splits), against 90-94% trained.
-Its confidence came out close to calibrated on those datasets (error 0.005-0.045 with 2 to 150 options), but check
+Zero-shot is for trying things out: 66% on BANKING77 and 73% on CLINC150 (dev splits), against 90-94% trained.
+Its confidence came out close to calibrated on those datasets (error 0.03-0.04), but check
 it on your own labels before relying on a threshold.
 
 ## Train a head
@@ -109,7 +109,7 @@ The [guide](docs/guide-typesafe.md) covers what differs, including what `confide
 
 ## switch
 
-`decidekit/switch` makes decidejs a decision model for [switch](https://github.com/osolmaz/switch), the harness that
+`decidekit/switch` makes decidekit a decision model for [switch](https://github.com/osolmaz/switch), the harness that
 runs a decision model through a task. Zero-shot, it won two of the three WikiRace races from switch's example on a
 laptop CPU. See the [switch guide](docs/guide-switch.md).
 
@@ -132,7 +132,7 @@ const d = await createWorkerDecider({
 const r = await d.decide(text, head);
 ```
 
-Import `loadHead` from `decidekit/browser` too, not from `decidejs`: that keeps the encoder out of your page bundle.
+Import `loadHead` from `decidekit/browser` too, not from `decidekit`: that keeps the encoder out of your page bundle.
 
 Tested from npm in Vite 7, webpack 5 and Next.js 16 apps; see [docs/bundlers.md](docs/bundlers.md) for the config, including a one-line alias
 that halves the wasm download. Serve the page cross-origin isolated if you can:
@@ -146,9 +146,24 @@ That lets wasm use several threads: in our runs with all-MiniLM-L6-v2 it cut p95
 
 Download on first visit: the model (34 MB) and onnxruntime's wasm (23.6 MB, 5.8 MB gzipped), both cached after.
 
+## Models
+
+decidekit runs the Decide models, each a sentence encoder plus a calibrated head:
+
+| Model | Id | Download | Reads | Upstream |
+|---|---|---|---|---|
+| Decide Small (default) | [`debuffjp/decide-small`](https://huggingface.co/debuffjp/decide-small) | 34 MB | English | BAAI's [bge-small-en-v1.5](https://huggingface.co/BAAI/bge-small-en-v1.5), identical to `Xenova/bge-small-en-v1.5` |
+| Decide Multilingual | [`debuffjp/decide-multilingual`](https://huggingface.co/debuffjp/decide-multilingual) | about 118 MB | 76% on Japanese and Chinese intents | intfloat's [multilingual-e5-small](https://huggingface.co/intfloat/multilingual-e5-small), identical to `Xenova/multilingual-e5-small` |
+
+Pass the id as `model` to `createDecider()`. The files download from Hugging Face on first use, pinned to a revision, and
+are then cached. To serve them yourself, pass `modelHost`: see [self-hosting](docs/guide-browser.md#host-the-model-files-yourself).
+
+Before 0.1.1 the default id was `Xenova/bge-small-en-v1.5`. The old ids and heads recording them keep working: they
+resolve to the new ids, so `d.model` returns the new id.
+
 ## Numbers
 
-bge-small-en-v1.5 (the default) at q8, Ryzen 7 260 laptop, Node 24 and headless Chrome 156, 2026-11-01. Accuracy
+Decide Small (bge-small-en-v1.5, the default) at q8, without label names, Ryzen 7 260 laptop, Node 24 and headless Chrome 156, 2026-11-01. Accuracy
 and calibration error (ECE, lower is better) on each dataset's test split, mean of 3 seeds (one run for "all"):
 [eval/results-bge-small.json](eval/results-bge-small.json).
 
@@ -185,16 +200,16 @@ in-scope examples only, with `minConfidence` 0.93 picked on the validation split
 
 **The smaller model.** `createDecider({ model: "Xenova/all-MiniLM-L6-v2" })` (23.7 MB, the pre-release default) is
 about 1.8x faster per decision in Chrome and 3 to 5 points less accurate with few examples, about the same with all
-the data. Pass it to keep using heads trained on it. Before this default changed, the README suggested
-`{ labelNames: false }` with bge-small: label names slightly hurt it on the dev splits.
+the data. Pass it to keep using heads trained on it. Before 0.1.0 a label name counted as 3 examples and the README
+suggested `{ labelNames: false }` with bge-small. It now counts as 1, which helps.
 
 Reproduce: `npm run eval:suite` (Node), `npm run bench` then `npm run bench:run` (Chrome).
 
 ## Playground
 
-Try it at [debuff.jp/decidejs](https://debuff.jp/decidejs/). To run it locally: `npm run playground`, then serve `playground/dist` cross-origin isolated (`node bench/serve.mjs` style headers).
+Try it at [debuff.jp/decide](https://debuff.jp/decide/). To run it locally: `npm run playground`, then serve `playground/dist` cross-origin isolated (`node bench/serve.mjs` style headers).
 Try zero-shot labels, the ready-made heads, a head trained in the page and "none of these".
-[Two lanes](https://debuff.jp/decidejs/lanes/) shows `withFallback` live: decidejs starts from twelve label names,
+[Two lanes](https://debuff.jp/decide/lanes/) shows `withFallback` live: decidekit starts from twelve label names,
 asks a small LLM when it's unsure, learns from each answer, and replays the first 20 messages at the end.
 
 ## Ready-made heads
@@ -213,30 +228,26 @@ and offline apps, and text that can't leave the device. Details, code and poor f
 
 | | Runs in the browser | Trains in the browser | Typed result | Calibrated confidence |
 |---|---|---|---|---|
-| decidejs | yes (worker) | yes | yes | trained heads |
+| decidekit | yes (worker) | yes | yes | trained heads |
 | [jimothy](https://www.npmjs.com/package/jimothy) | inference (worker) | no, Node | yes | yes |
 | [Jev](https://typesafe.ai/) (TypeSafe) | no, API | no | yes | yes |
 | transformers.js zero-shot pipeline | yes | no | no | no |
 | MediaPipe text classifier | yes | no, offline | no | no |
 | LLM structured output (AI SDK `Output.choice`) | no, API | n/a | yes | no |
 
-jimothy reports 92.4% on BANKING77, above our 90.7%. Pick it if you train in Node and want the higher number. Pick
-decidejs if training has to happen in the page, from the user's own data.
-
 ## Docs
 
-[docs/README.md](docs/README.md) indexes the guides (training, confidence and thresholds, the browser), the
+[debuff.jp/decide/docs](https://debuff.jp/decide/docs/) (also in [docs/README.md](docs/README.md)) indexes the guides (training, confidence and thresholds, the browser), the
 API reference, how it works and the FAQ.
 
 ## Community
 
-Questions and show-and-tell in [Discussions](https://github.com/debuffjp/decidejs/discussions), chat in the
+Questions and show-and-tell in [Discussions](https://github.com/debuffjp/decidekit/discussions), chat in the
 [debuff Discord](https://discord.gg/jshbDHSpCY). Results on your own data, good or bad, help most: there's an
 issue template for them. See [CONTRIBUTING.md](CONTRIBUTING.md).
 
 ## Status
 
-0.1. English with the default model, other languages with `multilingual-e5-small` ([FAQ](docs/faq.md#which-languages-does-it-support)); short texts (about a paragraph). Not done yet: WebGPU in the worker, a
-fallback to a hosted model below the threshold, and fine-tuning the encoder (the likely route past about 91%).
+0.1. English with the default model, other languages with Decide Multilingual ([FAQ](docs/faq.md#which-languages-does-it-support)); short texts (about a paragraph). Not done yet: WebGPU in the worker and fine-tuning the encoder (the likely route past about 91%).
 
 Made by the [debuff](https://debuff.jp) team. MIT.

package/dist/browser.js

@@ -36,7 +36,7 @@ export async function createWorkerDecider({ onProgress, ...options } = {}, worke
         terminate: () => {
             worker.terminate();
             for (const p of pending.values())
-                p.reject(new Error("decidejs: the worker was terminated"));
+                p.reject(new Error("decidekit: the worker was terminated"));
             pending.clear();
         },
     };

package/dist/fallback.js

@@ -25,9 +25,9 @@ export function withFallback(decider, using, options) {
         perLabel.set(e.label, (perLabel.get(e.label) ?? 0) + 1);
     function retrain() {
         if (!examples)
-            return Promise.reject(new Error("decidejs: withFallback() needs `examples` to retrain a head"));
+            return Promise.reject(new Error("decidekit: withFallback() needs `examples` to retrain a head"));
         if (!examples.length && !learned.length)
-            return Promise.reject(new Error("decidejs: nothing to train on yet: no answers from `ask`"));
+            return Promise.reject(new Error("decidekit: nothing to train on yet: no answers from `ask`"));
         const calibrateOn = trainOptions?.calibrateOn ?? (confident.length >= 50 ? confident : undefined);
         const trainLabels = learned.some((e) => e.label === NONE) ? [...labels, NONE] : labels;
         retraining ??= decider.train([...examples, ...learned], { ...trainOptions, labels: trainLabels, calibrateOn }).then((h) => {

package/dist/index.d.ts

@@ -72,13 +72,19 @@ export interface TrainOptions<L extends string = string> {
 }
 export interface DeciderOptions {
     /**
-     * A Hugging Face sentence-embedding model with ONNX weights. Default `Xenova/bge-small-en-v1.5` (34 MB at q8).
+     * A Hugging Face sentence-embedding model with ONNX weights. Default `debuffjp/decide-small` (Decide Small, 34 MB at q8).
+     * Before 0.1.1 the default id was `Xenova/bge-small-en-v1.5`, the same weights; that id and heads recording it still work.
      * Pre-release builds defaulted to `Xenova/all-MiniLM-L6-v2` (23 MB): smaller and 1.6x faster, 3 to 5 points less
      * accurate. Pass it to keep using heads trained on it.
      */
     model?: string;
     /** Defaults to transformers.js' choice: wasm in browsers, cpu in Node. */
     device?: "wasm" | "webgpu" | "cpu";
+    /**
+     * Serve model files from your own host instead of Hugging Face, as `<modelHost>/<model>/`, for example
+     * `https://cdn.example.com/models/debuffjp/decide-small/onnx/model_quantized.onnx`. Applies to the whole page or process.
+     */
+    modelHost?: string;
     /** Weight precision. Default `q8`. */
     dtype?: "q8" | "fp16" | "fp32";
     /** Called as model files download (first load only; later loads come from cache). */
@@ -98,7 +104,7 @@ export interface DeciderOptions {
  * `train()`. In a browser, prefer `createWorkerDecider()` from `decidekit/browser`: the encoder blocks the thread
  * it runs on.
  */
-export declare function createDecider({ model, device, dtype, temperature, onProgress }?: DeciderOptions): Promise<{
+export declare function createDecider({ model: requested, device, dtype, temperature, onProgress, modelHost }?: DeciderOptions): Promise<{
     model: string;
     decide: {
         <const L extends string>(text: string, using: Choices<L>, options?: {

package/dist/index.js

@@ -1,13 +1,13 @@
 // decide(text, options): typed, calibrated decisions on device. Zero-shot compares the text with each option's
 // embedding; a trained head (a softmax layer over the same embeddings, 156 KB for 77 labels) is far more
 // accurate. Both return the option literals as the choice type.
-import { pipeline } from "@huggingface/transformers";
+import { env, pipeline } from "@huggingface/transformers";
 import { fitHead, logits, softmax } from "./head.js";
 function isHead(c) {
     if (Array.isArray(c) || !("weights" in c))
         return false;
     if (!(c.weights instanceof Float32Array))
-        throw new Error("decidejs: this head's weights aren't a Float32Array. Read head files with loadHead(), not JSON.parse.");
+        throw new Error("decidekit: this head's weights aren't a Float32Array. Read head files with loadHead(), not JSON.parse.");
     return true;
 }
 const readable = (label) => label.replace(/[_-]+/g, " ").replace(/([a-z])([A-Z])/g, "$1 $2").toLowerCase();
@@ -23,9 +23,15 @@ const LABEL_NAME_WEIGHT = 1;
 // multilingual-e5-small's similarities sit close together, so it needs far lower temperatures: on MASSIVE in English
 // and Japanese, 0.0105 for label names (calibration error 0.03 to 0.04; descriptions not yet measured) and a floor of
 // 0.4 (eval/e5-settings.ts). It also expects "query: " before every text.
-const SMALL_HEAD_FLOOR = { "Xenova/bge-small-en-v1.5": 0.5, "Xenova/multilingual-e5-small": 0.4 };
-const TEMPERATURES = { "Xenova/bge-small-en-v1.5": [0.03, 0.025], "Xenova/all-MiniLM-L6-v2": [0.05, 0.05], "Xenova/multilingual-e5-small": [0.0105, 0.0105] };
-const PREFIX = { "Xenova/multilingual-e5-small": "query: " };
+// The Decide models are unmodified copies of these upstream ids, so the old ids and heads trained on them still work.
+const SMALL = "debuffjp/decide-small", MULTILINGUAL = "debuffjp/decide-multilingual";
+const SAME_MODEL = { "Xenova/bge-small-en-v1.5": SMALL, "Xenova/multilingual-e5-small": MULTILINGUAL };
+const canonical = (model) => SAME_MODEL[model] ?? model;
+// Pinned so a later change on Hugging Face can't change what an installed version decides.
+const REVISIONS = { [SMALL]: "188d690f86190863e3b4e44e29c88348b10230c4", [MULTILINGUAL]: "a3e353d26d846b2163db8acad65ef3d43893592e" };
+const SMALL_HEAD_FLOOR = { [SMALL]: 0.5, [MULTILINGUAL]: 0.4 };
+const TEMPERATURES = { [SMALL]: [0.03, 0.025], "Xenova/all-MiniLM-L6-v2": [0.05, 0.05], [MULTILINGUAL]: [0.0105, 0.0105] };
+const PREFIX = { [MULTILINGUAL]: "query: " };
 function toDecision(labels, p) {
     let best = 0;
     for (let i = 1; i < p.length; i++)
@@ -39,9 +45,14 @@ function toDecision(labels, p) {
  * `train()`. In a browser, prefer `createWorkerDecider()` from `decidekit/browser`: the encoder blocks the thread
  * it runs on.
  */
-export async function createDecider({ model = "Xenova/bge-small-en-v1.5", device, dtype = "q8", temperature, onProgress } = {}) {
+export async function createDecider({ model: requested = SMALL, device, dtype = "q8", temperature, onProgress, modelHost } = {}) {
+    const model = canonical(requested);
+    if (modelHost) {
+        env.remoteHost = modelHost.endsWith("/") ? modelHost : modelHost + "/";
+        env.remotePathTemplate = "{model}/";
+    }
     const embed = await pipeline("feature-extraction", model, {
-        device, dtype,
+        device, dtype, revision: REVISIONS[model],
         progress_callback: onProgress && ((p) => p.status === "progress" && onProgress({ file: p.file, loaded: p.loaded, total: p.total })),
     });
     const optionCache = new Map(), prefix = PREFIX[model] ?? "";
@@ -58,14 +69,14 @@ export async function createDecider({ model = "Xenova/bge-small-en-v1.5", device
     async function score(q, using, minConfidence) {
         let d;
         if (isHead(using)) {
-            if (using.model !== model)
-                throw new Error(`decidejs: this head was trained on ${using.model}, but the decider uses ${model}. Pass { model: "${using.model}" } to createDecider.`);
+            if (canonical(using.model) !== model)
+                throw new Error(`decidekit: this head was trained on ${using.model}, but the decider uses ${model}. Pass { model: "${using.model}" } to createDecider.`);
             d = toDecision(using.labels, softmax(logits(using, q), using.temperature));
         }
         else {
             const labels = (Array.isArray(using) ? using : Object.keys(using));
             if (labels.length < 2)
-                throw new Error(`decidejs: decide() needs at least 2 options, got ${labels.length}`);
+                throw new Error(`decidekit: decide() needs at least 2 options, got ${labels.length}`);
             // Bare labels are compared as readable text ("card_arrival" as "card arrival"): 2.6 points better on CLINC150's
             // dev split. A described label scores by the better of its name and its description: on the dev splits that
             // beat the description alone by 0.5 (BANKING77) and 3.1 points (CLINC150), and names alone by 7 and 10.
@@ -98,15 +109,15 @@ export async function createDecider({ model = "Xenova/bge-small-en-v1.5", device
     async function train(examples, opts = {}) {
         const bad = examples.findIndex((e) => typeof e?.text !== "string" || typeof e?.label !== "string" || !e.label);
         if (bad >= 0)
-            throw new Error(`decidejs: example ${bad} needs a string text and a non-empty string label, got ${JSON.stringify(examples[bad])}`);
+            throw new Error(`decidekit: example ${bad} needs a string text and a non-empty string label, got ${JSON.stringify(examples[bad])}`);
         if (!examples.length)
-            throw new Error("decidejs: train() needs at least 1 example");
+            throw new Error("decidekit: train() needs at least 1 example");
         const labels = [...new Set([...(opts.labels ?? []), ...examples.map((e) => e.label)])].sort();
         if (opts.labels && opts.labelNames === false && labels.some((l) => !examples.some((e) => e.label === l))) {
-            throw new Error("decidejs: labels without examples need labelNames, which is what they start from");
+            throw new Error("decidekit: labels without examples need labelNames, which is what they start from");
         }
         if (labels.length < 2)
-            throw new Error(`decidejs: train() needs at least 2 labels, got ${labels.length}`);
+            throw new Error(`decidekit: train() needs at least 2 labels, got ${labels.length}`);
         const y = examples.map((e) => labels.indexOf(e.label));
         const texts = [...new Set(examples.map((e) => e.text))];
         const fresh = texts.filter((t) => !trainCache.has(t));
@@ -122,7 +133,7 @@ export async function createDecider({ model = "Xenova/bge-small-en-v1.5", device
         const cal = opts.calibrateOn;
         const unknown = cal?.find((e) => !labels.includes(e.label));
         if (unknown)
-            throw new Error(`decidejs: calibrateOn has label ${JSON.stringify(unknown.label)}, which isn't one of the head's labels`);
+            throw new Error(`decidekit: calibrateOn has label ${JSON.stringify(unknown.label)}, which isn't one of the head's labels`);
         const calibration = cal?.length ? { X: await vectors(cal.map((e) => e.text)), y: cal.map((e) => labels.indexOf(e.label)) } : undefined;
         return { model, ...(await fitHead(X, y, labels, opts, priors, LABEL_NAME_WEIGHT, undefined, SMALL_HEAD_FLOOR[model] ?? 1, calibration)) };
     }

package/dist/switch.d.ts

@@ -81,22 +81,22 @@ export interface SwitchOptions {
     /** Trained heads for questions that repeat, by question id. Questions without one are decided zero-shot. */
     heads?: Readonly<Record<string, Head>>;
     /**
-     * A stronger decider to ask when decidejs isn't sure, such as Jev through switch's pi-ai decider. Its answers to
+     * A stronger decider to ask when decidekit isn't sure, such as Jev through switch's pi-ai decider. Its answers to
      * choice questions train a head for that question, so a question that repeats asks it less and less (see `withFallback`).
      */
     teacher?: SwitchDecider;
     /** Below this confidence, choice questions go to `teacher`. Default 0.9. */
     minConfidence?: number;
     /**
-     * What decidejs reads for a question. Its encoder reads about 256 tokens and drops the rest, so a long state or long
+     * What decidekit reads for a question. Its encoder reads about 256 tokens and drops the rest, so a long state or long
      * instructions can push out what matters; pick it here, such as the target of a search. Default: `textOf`.
      */
     text?: (state: SwitchRequest["decisionState"], question: Question, id: string) => string;
 }
-/** The text decidejs reads for a question: its instructions, then the state as plain lines rather than JSON. */
+/** The text decidekit reads for a question: its instructions, then the state as plain lines rather than JSON. */
 export declare function textOf(instructions: string, state: Json): string;
 /**
- * Wraps a decidejs decider as a switch decider. Choice, bool and score questions become decidejs decisions over the
+ * Wraps a decidekit decider as a switch decider. Choice, bool and score questions become decidekit decisions over the
  * question's criteria; images are ignored. Every call is local, so usage costs nothing.
  */
 export declare function switchDecider(local: Local, options?: SwitchOptions): SwitchDecider;

package/dist/switch.js

@@ -1,8 +1,8 @@
-// decidejs as a decision model for switch (github.com/osolmaz/switch), a harness that runs a decision model through a
-// task by asking it typed questions. The types below copy the parts of switch's `Decider` that this needs, so decidejs
+// decidekit as a decision model for switch (github.com/osolmaz/switch), a harness that runs a decision model through a
+// task by asking it typed questions. The types below copy the parts of switch's `Decider` that this needs, so decidekit
 // doesn't depend on switch: pass the result wherever switch takes a decider.
 import { withFallback } from "./fallback.js";
-/** The text decidejs reads for a question: its instructions, then the state as plain lines rather than JSON. */
+/** The text decidekit reads for a question: its instructions, then the state as plain lines rather than JSON. */
 export function textOf(instructions, state) {
     const lines = [];
     const walk = (v, key) => {
@@ -18,7 +18,7 @@ export function textOf(instructions, state) {
     return [instructions, ...lines].filter(Boolean).join("\n");
 }
 /**
- * Wraps a decidejs decider as a switch decider. Choice, bool and score questions become decidejs decisions over the
+ * Wraps a decidekit decider as a switch decider. Choice, bool and score questions become decidekit decisions over the
  * question's criteria; images are ignored. Every call is local, so usage costs nothing.
  */
 export function switchDecider(local, options = {}) {
@@ -48,7 +48,7 @@ export function switchDecider(local, options = {}) {
         const text = textFor(state, q, id);
         if (q.type === "choice") {
             const keys = Object.keys(q.criteria);
-            // decidejs compares two or more options; a single one is the answer.
+            // decidekit compares two or more options; a single one is the answer.
             if (keys.length === 1)
                 return { type: "choice", choice: keys[0], probabilities: { [keys[0]]: 1 }, confidence: 1 };
             if (teacher)
@@ -66,14 +66,14 @@ export function switchDecider(local, options = {}) {
         return { type: "score", score: Number(d.choice), confidence: d.confidence };
     }
     return {
-        model: `decidejs/${local.model}`,
+        model: `decidekit/${local.model}`,
         // The encoder reads about 256 tokens and cuts the rest, so a long state still fits; tell switch the window is big
         // so it doesn't trim the choices to fit.
         window: 1_000_000,
         charsPerToken: 4,
         async decide(request) {
             const started = performance.now();
-            const base = { api: "decidejs", provider: "decidejs", model: `decidejs/${local.model}`, timestamp: Date.now(), tooLong: false };
+            const base = { api: "decidekit", provider: "decidekit", model: `decidekit/${local.model}`, timestamp: Date.now(), tooLong: false };
             try {
                 const entries = await Promise.all(Object.entries(request.decisionQuestions).map(async ([id, q]) => [id, await answer(id, q, request.decisionState)]));
                 const input = Math.round(JSON.stringify(request.decisionState).length / 4);

package/dist/typesafe.d.ts

@@ -5,7 +5,7 @@ export interface Decides {
 }
 export interface ChoiceQuestion<L extends string = string> {
     type: "choice";
-    /** Kept for compatibility and ignored: decidejs compares the state with the options, it doesn't read a question. */
+    /** Kept for compatibility and ignored: decidekit compares the state with the options, it doesn't read a question. */
     instructions: string;
     /** Option names to descriptions (`null` for none), as in TypeSafe's API, or a trained head. */
     criteria: Readonly<Record<L, string | null>> | Head<L>;

package/dist/typesafe.js

@@ -2,7 +2,7 @@ export function choice(instructions, criteria) {
     return { type: "choice", instructions, criteria };
 }
 const unsupported = (type) => () => {
-    throw new Error(`decidejs: ${type} questions aren't supported locally, only choice. Ask a yes/no as a choice between two described options.`);
+    throw new Error(`decidekit: ${type} questions aren't supported locally, only choice. Ask a yes/no as a choice between two described options.`);
 };
 export const noul = unsupported("noul");
 export const score = unsupported("score");
@@ -15,7 +15,7 @@ function stateText(state) {
             return v;
         if (typeof v === "number" || typeof v === "boolean")
             return String(v);
-        throw new Error("decidejs: state values must be text, numbers or booleans");
+        throw new Error("decidekit: state values must be text, numbers or booleans");
     }).join("\n");
 }
 function options(criteria) {
@@ -41,7 +41,7 @@ export function typesafeClient(decider) {
                 const n = Object.keys(probabilities).length;
                 answers[id] = { type: "choice", choice, confidence: Math.max(0, (n * confidence - 1) / (n - 1)), probabilities };
             }
-            return { model: "decidejs", answers: answers };
+            return { model: "decidekit", answers: answers };
         },
     };
 }

package/package.json

@@ -1,93 +1,93 @@
-{
-  "name": "decidekit",
-  "version": "0.0.1",
-  "description": "A System One model in the browser: typed, calibrated text decisions on the user's device. Learns from an LLM when it isn't sure.",
-  "license": "MIT",
-  "repository": "github:debuffjp/decidejs",
-  "keywords": [
-    "decision model",
-    "system one",
-    "classification",
-    "intent",
-    "zero-shot",
-    "calibration",
-    "llm",
-    "embeddings",
-    "browser",
-    "typescript",
-    "text classification",
-    "transformers.js",
-    "on-device",
-    "llm fallback"
-  ],
-  "type": "module",
-  "exports": {
-    ".": {
-      "types": "./dist/index.d.ts",
-      "default": "./dist/index.js"
-    },
-    "./browser": {
-      "types": "./dist/browser.d.ts",
-      "default": "./dist/browser.js"
-    },
-    "./react": {
-      "types": "./dist/react.d.ts",
-      "default": "./dist/react.js"
-    },
-    "./switch": {
-      "types": "./dist/switch.d.ts",
-      "default": "./dist/switch.js"
-    },
-    "./typesafe": {
-      "types": "./dist/typesafe.d.ts",
-      "default": "./dist/typesafe.js"
-    }
-  },
-  "files": [
-    "dist",
-    "CHANGELOG.md"
-  ],
-  "scripts": {
-    "build": "tsc -p tsconfig.build.json",
-    "prepack": "npm run build",
-    "typecheck": "tsc --noEmit",
-    "eval": "node --experimental-strip-types eval/banking77.ts",
-    "bench": "esbuild src/worker.ts src/browser.ts --bundle --format=esm --platform=browser --outdir=bench --alias:onnxruntime-web/webgpu=onnxruntime-web/wasm && node bench/serve.mjs",
-    "playground": "node playground/build.mjs",
-    "heads": "node --experimental-strip-types heads/build.ts",
-    "bench:run": "node bench/run.mjs",
-    "eval:suite": "node --experimental-strip-types eval/suite.ts",
-    "eval:hill": "node --experimental-strip-types eval/hill.ts",
-    "test": "node --test --experimental-strip-types \"test/*.test.ts\"",
-    "lint:prose": "node scripts/lint-prose.mjs",
-    "smoke": "node scripts/smoke.mjs"
-  },
-  "dependencies": {
-    "@huggingface/transformers": "4.2.0"
-  },
-  "devDependencies": {
-    "@types/node": "22.15.3",
-    "@types/react": "19.3.0",
-    "esbuild": "0.25.10",
-    "playwright-core": "1.64.1",
-    "react": "19.3.1",
-    "spryte": "0.11.0",
-    "typescript": "5.9.3"
-  },
-  "engines": {
-    "node": ">=22.18"
-  },
-  "homepage": "https://debuff.jp/decidejs/",
-  "bugs": "https://github.com/debuffjp/decidejs/issues",
-  "bin": {
-    "decidekit": "./dist/cli.js"
-  },
-  "peerDependencies": {
-    "react": ">=18"
-  },
-  "peerDependenciesMeta": {
-    "react": {
-      "optional": true
-    }
-  }
-}
+{
+  "name": "decidekit",
+  "version": "0.1.1",
+  "description": "A System One model in the browser: typed, calibrated text decisions on the user's device. Learns from an LLM when it isn't sure.",
+  "license": "MIT",
+  "repository": "github:debuffjp/decidekit",
+  "keywords": [
+    "decision model",
+    "system one",
+    "classification",
+    "intent",
+    "zero-shot",
+    "calibration",
+    "llm",
+    "embeddings",
+    "browser",
+    "typescript",
+    "text classification",
+    "transformers.js",
+    "on-device",
+    "llm fallback"
+  ],
+  "type": "module",
+  "exports": {
+    ".": {
+      "types": "./dist/index.d.ts",
+      "default": "./dist/index.js"
+    },
+    "./browser": {
+      "types": "./dist/browser.d.ts",
+      "default": "./dist/browser.js"
+    },
+    "./react": {
+      "types": "./dist/react.d.ts",
+      "default": "./dist/react.js"
+    },
+    "./switch": {
+      "types": "./dist/switch.d.ts",
+      "default": "./dist/switch.js"
+    },
+    "./typesafe": {
+      "types": "./dist/typesafe.d.ts",
+      "default": "./dist/typesafe.js"
+    }
+  },
+  "files": [
+    "dist",
+    "CHANGELOG.md"
+  ],
+  "scripts": {
+    "build": "tsc -p tsconfig.build.json",
+    "prepack": "npm run build",
+    "typecheck": "tsc --noEmit",
+    "eval": "node --experimental-strip-types eval/banking77.ts",
+    "bench": "esbuild src/worker.ts src/browser.ts --bundle --format=esm --platform=browser --outdir=bench --alias:onnxruntime-web/webgpu=onnxruntime-web/wasm && node bench/serve.mjs",
+    "playground": "node playground/build.mjs",
+    "heads": "node --experimental-strip-types heads/build.ts",
+    "bench:run": "node bench/run.mjs",
+    "eval:suite": "node --experimental-strip-types eval/suite.ts",
+    "eval:hill": "node --experimental-strip-types eval/hill.ts",
+    "test": "node --test --experimental-strip-types \"test/*.test.ts\"",
+    "lint:prose": "node scripts/lint-prose.mjs",
+    "smoke": "node scripts/smoke.mjs"
+  },
+  "dependencies": {
+    "@huggingface/transformers": "4.2.0"
+  },
+  "devDependencies": {
+    "@types/node": "22.15.3",
+    "@types/react": "19.3.0",
+    "esbuild": "0.25.10",
+    "playwright-core": "1.64.1",
+    "react": "19.3.1",
+    "spryte": "0.11.0",
+    "typescript": "5.9.3"
+  },
+  "engines": {
+    "node": ">=22.18"
+  },
+  "homepage": "https://debuff.jp/decide/",
+  "bugs": "https://github.com/debuffjp/decidekit/issues",
+  "bin": {
+    "decidekit": "./dist/cli.js"
+  },
+  "peerDependencies": {
+    "react": ">=18"
+  },
+  "peerDependenciesMeta": {
+    "react": {
+      "optional": true
+    }
+  }
+}