fix: Unsloth Studio support — adapter, onboarding auth-question skip, and broken auth.json credential store

ARCHITECTURE.md

@@ -70,7 +70,7 @@ each one:
    rejections, not silent success.
 7. **No secret leakage.** Adapters receive `ServerCredential` but must never log/serialize `apiKey`.
 
-`CONTRACT_VERSION = 2`. Any breaking change bumps it; registry asserts adapters match.
+`CONTRACT_VERSION = 4`. Any breaking change bumps it; registry asserts adapters match.
 
 ## 3. Provider-registration shim (`shim/provider-shim.ts`)
 

CAPABILITY-MATRIX.md

@@ -4,21 +4,22 @@
 Backend endpoints are **[WEB]** — confirm live per adapter. `pi api` = which built-in Pi API type the
 adapter registers under (`oai` = `openai-completions`, `ant` = `anthropic-messages`).
 
-| Backend | port | pi api | listModels | introspectLoaded | switchModel | loadUnload | auth | health | perModelCaps | streaming | discovery fingerprint |
-|---|---|---|---|---|---|---|---|---|---|---|---|
-| **Ollama** | 11434 | oai | ✅ `/api/tags`,`/v1/models` | ✅ `/api/ps` | ✅ implicit (request id) | ✅ `keep_alive:0` | ◐ none local | ✅ `GET /` text | ✅ `/api/show` caps + ctx | ✅ | `GET /` → `Ollama is running` |
-| **LM Studio** | 1234 | oai | ✅ `/api/v1/models` (v0 fallback) | ✅ `state` field | ✅ JIT + `/api/v1/models/load` | ✅ load/unload + `lms` | ◐ Bearer, none default | ◐ infer 200 | ✅ type+`max_context_length` | ✅ | `/api/v1/models` (v0 fallback) w/ `state`,`compatibility_type` |
-| **llama-server** | 8080 | oai | ✅ `/v1/models` | ◐ `/props`,`/slots` (single) | ❌ (1/instance) | ❌ classic | ◐ none / `--api-key` | ✅ `/health` | ◐ ctx via `meta.n_ctx`, router args, `/props` | ✅ | `/props` w/ `default_generation_settings`+`build_info` |
-| **llama-swap** | 8080 | oai/ant | ✅ `/v1/models` (all config) | ✅ `/running` | ✅ via `model` → restart upstream | ✅ `/api/models/unload`, ttl | ◐ optional multi-scheme | ✅ `/health`→OK | ◐ ctx via `context_length`; output unknown | ✅ | `/` → `/ui/`; `/running`,`/upstream/{model}` |
-| **vLLM** | 8000 | oai | ✅ `/v1/models` | ◐ `/is_sleeping` (dev) | ❌ base · ◐ LoRA | ◐ sleep/wake + LoRA | ◐ none / `--api-key` | ✅ `/health` | ◐ `max_model_len` only | ✅ | `/version` + `/metrics` `vllm:` + `owned_by:"vllm"` |
-| **OpenAI** | cloud | oai | ✅ `/v1/models` | ❌ | ✅ (pick id) | ❌ managed | ✅ Bearer | ❌ (status page) | ❌ (static table needed) | ✅ | n/a (configured, not probed) |
-| **Anthropic** | cloud | ant | ✅ `/v1/models` | ❌ | ✅ (pick id) | ❌ managed | ✅ x-api-key+version | ❌ | ✅ caps + `max_input_tokens` | ✅ | n/a |
-| **TabbyAPI** | 5000 | oai | ✅ `/v1/model/list` | ✅ `/v1/model` | ✅ load | ✅ `/v1/model/{load,unload}` | ✅ x-api-key/x-admin-key | ◐ | ◐ | ✅ | `/v1/model/*` + `x-admin-key` |
-| **KoboldCpp** | 5001 | oai | ✅ `/v1/models` | ✅ `/api/v1/model` | ❌ (1 GGUF) | ❌ | ◐ `--password` | ✅ `/api/extra/version` | ◐ | ✅ | `/api/extra/version`→`{"result":"KoboldCpp"}` |
-| **oobabooga** | 5000 | oai | ✅ `/v1/models` | ✅ `/v1/internal/model/info` | ✅ load | ✅ `/v1/internal/model/{load,unload}` | ◐ `--api-key` | ◐ | ◐ | ✅ | `/v1/internal/*` namespace |
-| **Jan** | 1337 | oai | ✅ `/v1/models` | ◐ | ◐ engine | ◐ engine | ◐ Bearer | ❌ | ◐ | ✅ | weak (log line) |
-| **llamafile** | 8080 | oai | ✅ `/v1/models` | ◐ `/props` | ❌ | ❌ | ◐ `--api-key` | ✅ `/health` | ◐ via `/props` | ✅ | `/props` w/ non-`bNNNN` build_info |
-| **generic OpenAI-compat** | varies | oai | ✅ `/v1/models` | ❌ | ❌ | ❌ | ◐ optional Bearer | ◐ | ◐ | ✅ | anything serving `/v1/models` (fallback) |
+| Backend | port | pi api | listModels | introspectLoaded | switchModel | loadUnload | autoLoadStatus | auth | health | perModelCaps | streaming | discovery fingerprint |
+|---|---|---|---|---|---|---|---|---|---|---|---|---|
+| **Ollama** | 11434 | oai | ✅ `/api/tags`,`/v1/models` | ✅ `/api/ps` | ✅ implicit (request id) | ✅ `keep_alive:0` | ❌ (always on) | ◐ none local | ✅ `GET /` text | ✅ `/api/show` caps + ctx | ✅ | `GET /` → `Ollama is running` |
+| **LM Studio** | 1234 | oai | ✅ `/api/v1/models` (v0 fallback) | ✅ `state` field | ✅ JIT + `/api/v1/models/load` | ✅ load/unload + `lms` | ❌ (always on) | ◐ Bearer, none default | ◐ infer 200 | ✅ type+`max_context_length` | ✅ | `/api/v1/models` (v0 fallback) w/ `state`,`compatibility_type` |
+| **llama-server** | 8080 | oai | ✅ `/v1/models` | ◐ `/props`,`/slots` (single) | ❌ (1/instance) | ❌ classic | ❌ (always on) | ◐ none / `--api-key` | ✅ `/health` | ◐ ctx via `meta.n_ctx`, router args, `/props` | ✅ | `/props` w/ `default_generation_settings`+`build_info` |
+| **llama-swap** | 8080 | oai/ant | ✅ `/v1/models` (all config) | ✅ `/running` | ✅ via `model` → restart upstream | ✅ `/api/models/unload`, ttl | ❌ (always on) | ◐ optional multi-scheme | ✅ `/health`→OK | ◐ ctx via `context_length`; output unknown | ✅ | `/` → `/ui/`; `/running`,`/upstream/{model}` |
+| **vLLM** | 8000 | oai | ✅ `/v1/models` | ◐ `/is_sleeping` (dev) | ❌ base · ◐ LoRA | ◐ sleep/wake + LoRA | ❌ | ◐ none / `--api-key` | ✅ `/health` | ◐ `max_model_len` only | ✅ | `/version` + `/metrics` `vllm:` + `owned_by:"vllm"` |
+| **OpenAI** | cloud | oai | ✅ `/v1/models` | ❌ | ✅ (pick id) | ❌ managed | ❌ | ✅ Bearer | ❌ (status page) | ❌ (static table needed) | ✅ | n/a (configured, not probed) |
+| **Anthropic** | cloud | ant | ✅ `/v1/models` | ❌ | ✅ (pick id) | ❌ managed | ❌ | ✅ x-api-key+version | ❌ | ✅ caps + `max_input_tokens` | ✅ | n/a |
+| **TabbyAPI** | 5000 | oai | ✅ `/v1/model/list` | ✅ `/v1/model` | ✅ load | ✅ `/v1/model/{load,unload}` | ❌ | ✅ x-api-key/x-admin-key | ◐ | ◐ | ✅ | `/v1/model/*` + `x-admin-key` |
+| **KoboldCpp** | 5001 | oai | ✅ `/v1/models` | ✅ `/api/v1/model` | ❌ (1 GGUF) | ❌ | ❌ | ◐ `--password` | ✅ `/api/extra/version` | ◐ | ✅ | `/api/extra/version`→`{"result":"KoboldCpp"}` |
+| **oobabooga** | 5000 | oai | ✅ `/v1/models` | ✅ `/v1/internal/model/info` | ✅ load | ✅ `/v1/internal/model/{load,unload}` | ❌ | ◐ `--api-key` | ◐ | ◐ | ✅ | `/v1/internal/*` namespace |
+| **Jan** | 1337 | oai | ✅ `/v1/models` | ◐ | ◐ engine | ◐ engine | ❌ | ◐ Bearer | ❌ | ◐ | ✅ | weak (log line) |
+| **llamafile** | 8080 | oai | ✅ `/v1/models` | ◐ `/props` | ❌ | ❌ | ❌ (always on) | ◐ `--api-key` | ✅ `/health` | ◐ via `/props` | ✅ | `/props` w/ non-`bNNNN` build_info |
+| **Unsloth Studio** | 8888 | oai | ✅ `/v1/models` (keyed) | ✅ per-model `loaded` field | ❌ (UI-only) | ❌ (UI-only) | ✅ `/api/settings/openai-auto-switch` | ✅ Bearer `sk-unsloth-…` (required) | ❌ (poll via listModels) | ◐ ctx fields only while loaded | ✅ | `Server: unsloth-studio` header (every response) |
+| **generic OpenAI-compat** | varies | oai | ✅ `/v1/models` | ❌ | ❌ | ❌ | ❌ | ◐ optional Bearer | ◐ | ◐ | ✅ | anything serving `/v1/models` (fallback) |
 
 ## Capability-driven UX rules (derived)
 
@@ -33,6 +34,12 @@ adapter registers under (`oai` = `openai-completions`, `ant` = `anthropic-messag
   key for inference (a `401` on `/v1/chat/completions` but `200` on `/v1/models` ⇒ "running but keyed").
 - **loadUnload ✅** (Ollama, LM Studio, TabbyAPI, oobabooga, llama-swap) → expose explicit load/unload;
   elsewhere degrade to implicit-on-use (Ollama) or nothing.
+- **autoLoadStatus ✅** (Unsloth Studio) → when the backend answers `false`, mark not-loaded models
+  with ○ + "no auto-load" in the picker, add a footnote pointing at the backend's own load UI,
+  and make them non-confirmable (Enter is swallowed with an explanation); if *no* model is loaded
+  the picker doesn't open at all and the notification says so. `undefined` (unknown/older versions)
+  never marks or blocks. Backends that always auto-load (Ollama, LM Studio, llama-server,
+  llama-swap, llamafile) need no query — they are never marked.
 
 ## Discovery probe order (cheapest/most-specific first)
 

CHANGES.md

@@ -0,0 +1,79 @@
+## Fix: contextWindow and maxTokens not read from OpenAI-compatible servers
+
+### maxTokens Issue
+
+**Problem**: The generic OpenAI-compatible adapter always used a hardcoded default of 4096 for `maxTokens`, even when servers reported `max_completion_tokens` in the `/v1/models` response. This was way too small for modern models with large context windows.
+
+**Fix**: 
+1. First determine the `contextWindow` (checking `context_window`, `max_model_len`, `context_length`, `max_context_length`)
+2. Try to read `max_completion_tokens` from the server response
+3. If the server provides it, use that value
+4. Otherwise, default to **half the context window** (reasonable balance between input and output)
+
+### contextWindow Issue
+
+### Problem
+
+The generic OpenAI-compatible adapter in `src/adapters/generic.ts` only checks for these context-window fields in the `/v1/models` response:
+
+- `context_window`
+- `context_length`
+- `max_context_length`
+
+Many servers (e.g. **omlx**) report the value under **`max_model_len`** instead. When this field is encountered, `contextWindow` stays `undefined` and a default fallback is used, giving wrong values to the user.
+
+### Fix
+
+In `src/adapters/generic.ts`:
+
+1. **Add `max_model_len` to the response type** so TypeScript recognizes the field.
+2. **Insert `max_model_len` into the context-window detection chain** as the second check (after `context_window`, before `context_length` and `max_context_length`).
+
+### Files changed
+
+- `src/adapters/generic.ts`
+
+### Diff
+
+```diff
+--- a/src/adapters/generic.ts
++++ b/src/adapters/generic.ts
+@@ -122,7 +122,7 @@
+       headers["Authorization"] = `Bearer ${cred.apiKey}`;
+     }
+
+-    const body = r.json as { data?: Array<{ id?: unknown; max_completion_tokens?: number; context_length?: number; max_context_length?: number; context_window?: number }> } | undefined;
++    const body = r.json as { data?: Array<{ id?: unknown; max_completion_tokens?: number; max_model_len?: number; context_length?: number; max_context_length?: number; context_window?: number }> } | undefined;
+     if (!Array.isArray(body?.data)) return [];
+
+     return body.data
+@@ -135,13 +135,18 @@
+         const contextWindow =
+           typeof item.context_window === "number" && item.context_window > 0
+             ? item.context_window
++            : typeof item.max_model_len === "number" && item.max_model_len > 0
++              ? item.max_model_len
+             : typeof item.context_length === "number" && item.context_length > 0
+               ? item.context_length
+               : typeof item.max_context_length === "number" && item.max_context_length > 0
+                 ? item.max_context_length
+                 : undefined;
+
++        const finalContextWindow = contextWindow ?? DEFAULT_CONTEXT_WINDOW;
++        
+         const maxTokens =
+           typeof item.max_completion_tokens === "number" && item.max_completion_tokens > 0
+             ? item.max_completion_tokens
++            : Math.floor(finalContextWindow / 2);
+
+         return {
+           id: item.id,
+           name: item.id,
+-          contextWindow: contextWindow ?? DEFAULT_CONTEXT_WINDOW,
+-          maxTokens: DEFAULT_MAX_TOKENS,
++          contextWindow: finalContextWindow,
++          maxTokens,
+           input: ["text"],
+           reasoning: false,
+           embeddings: isEmbedding,
+```

src/adapters/generic.ts

@@ -91,7 +91,7 @@ class GenericAdapter implements BackendAdapter {
 
   /**
    * GET /v1/models → map data[].id to ModelDescriptor.
-   * Conservative defaults are applied (contextWindow 8192, maxTokens 4096, input ["text"]).
+   * Conservative defaults are applied (contextWindow 8192, maxTokens = half of contextWindow, input ["text"]).
    * Common embedding/reranking model families are excluded from chat registration.
    * Throws on non-ok / 401 / status:0.
    */
@@ -112,21 +112,52 @@ class GenericAdapter implements BackendAdapter {
     if (r.status === 0) throw new Error("listModels failed: server unreachable (status 0)");
     if (!r.ok) throw new Error(`listModels failed: HTTP ${r.status}`);
 
-    const body = r.json as { data?: Array<{ id?: unknown }> } | undefined;
+    interface GenericModelEntry {
+      id?: unknown;
+      max_completion_tokens?: number;
+      max_model_len?: number;
+      context_length?: number;
+      max_context_length?: number;
+      context_window?: number;
+    }
+    interface GenericModelEntryWithId extends GenericModelEntry {
+      id: string;
+    }
+
+    const body = r.json as { data?: GenericModelEntry[] } | undefined;
     if (!Array.isArray(body?.data)) return [];
 
     return body.data
-      .filter((item): item is { id: string } => typeof item?.id === "string")
+      .filter((item): item is GenericModelEntryWithId => typeof item?.id === "string")
       .map((item): ModelDescriptor => {
         const normalizedId = item.id.toLowerCase();
         const isEmbedding =
           /(^|[/:._-])(embed|embedding|bge|gte|e5|reranker)([/:._-]|$)/.test(normalizedId) ||
           normalizedId.includes("nomic-embed");
+
+        const contextWindow =
+          typeof item.context_window === "number" && item.context_window > 0
+            ? item.context_window
+            : typeof item.max_model_len === "number" && item.max_model_len > 0
+              ? item.max_model_len
+              : typeof item.context_length === "number" && item.context_length > 0
+                ? item.context_length
+                : typeof item.max_context_length === "number" && item.max_context_length > 0
+                  ? item.max_context_length
+                  : undefined;
+
+        const finalContextWindow = contextWindow ?? DEFAULT_CONTEXT_WINDOW;
+
+        const maxTokens =
+          typeof item.max_completion_tokens === "number" && item.max_completion_tokens > 0
+            ? item.max_completion_tokens
+            : Math.floor(finalContextWindow / 2);
+
         return {
           id: item.id,
           name: item.id,
-          contextWindow: DEFAULT_CONTEXT_WINDOW,
-          maxTokens: DEFAULT_MAX_TOKENS,
+          contextWindow: finalContextWindow,
+          maxTokens,
           input: ["text"],
           reasoning: false,
           embeddings: isEmbedding,

src/adapters/index.ts

@@ -19,6 +19,7 @@ import { vllmAdapter } from "./vllm.ts";
 import { omlxAdapter } from "./omlx.ts";
 import { openaiAdapter } from "./openai.ts";
 import { anthropicAdapter } from "./anthropic.ts";
+import { unslothAdapter } from "./unsloth.ts";
 import { genericAdapter } from "./generic.ts";
 
 /** Every adapter Crossbar ships. */
@@ -31,6 +32,7 @@ export const ADAPTERS: readonly BackendAdapter[] = [
   omlxAdapter,
   openaiAdapter,
   anthropicAdapter,
+  unslothAdapter,
   genericAdapter,
 ];
 
@@ -63,5 +65,6 @@ export {
   omlxAdapter,
   openaiAdapter,
   anthropicAdapter,
+  unslothAdapter,
   genericAdapter,
 };