@billjr99/pi-openai-compat 1.1.17 → 1.1.18

package/README.md

@@ -59,26 +59,6 @@ If pi is already running when you install, type `/reload` first.
 | **llmproxy** | `http://localhost:8080/v1` (editable) | Keyless by default; bearer token if your instance requires one |
 | **Custom** | Any URL you supply | Optional bearer token |
 
-> **The `__` → `/` model-id rewrite (display compatibility).**
-> Some providers advertise model ids in a `provider__model` form. For example,
-> [llmproxy](https://github.com/BillJr99/llmproxy) exposes ids like
-> `openrouter__gpt-4` and virtual models such as `llmproxy__free` and
-> `llmproxy__loadbalanced`. The extension can optionally rewrite the first `__`
-> of each id to `/` (e.g. `openrouter__gpt-4` → `openrouter/gpt-4`,
-> `llmproxy__free` → `llmproxy/free`) before registering with pi, so the ids
-> display in the more familiar `provider/model` slash form. Upstreams that use
-> the `__` convention canonicalize the slash form back to `__` on each request,
-> so routing still works.
->
-> This is a presentation-only convenience controlled by a per-provider config
-> field (`rewriteDoubleUnderscore`, default **false**). It is enabled
-> automatically for the llmproxy template purely for display consistency — other
-> providers are unaffected, and you can turn it on or off freely. If you added a
-> provider as a **Custom** endpoint and want the slash-form display, set
-> `"rewriteDoubleUnderscore": true` on that provider in
-> `~/.config/pi-openai-compat/config.json` (or re-run `/compat-login` and pick a
-> template that enables it). Leaving it off keeps the original `__` ids.
-
 > **Providers whose model catalog lives at a non-standard `/models` path (as of June 2026)**
 > Some providers don't return models at `<base_url>/models`. The extension
 > handles them in one of two ways:
@@ -353,9 +333,6 @@ The usual cause is a **scoped or restricted API key/account** — the catalog
 endpoint advertises the full model list, but your credentials are only entitled
 to a subset, so pi only surfaces the ones you can actually use. Check whether your
 key is limited to specific models or a free tier on the provider's dashboard.
-(This is unrelated to the `__` → `/` rewrite, which is a display-only convenience —
-see the note in **Supported providers** above. Toggling `rewriteDoubleUnderscore`
-only changes how ids are displayed, not how many appear.)
 
 **Models appear in `/model` but requests fail**
 Check `/compat-login` ran successfully (no error message).

package/index.ts

@@ -48,13 +48,6 @@ interface ProviderConfig {
   // matches its template and can't be auto-migrated, so the notice is shown
   // only once rather than on every session_start while the config stays stale.
   staleNotified?: boolean;
-  // When true, rewrite the FIRST "__" of every model id to "/" before
-  // registering with pi, so ids from providers that use a "provider__model"
-  // convention (e.g. llmproxy) display in the more familiar "provider/model"
-  // slash form. This is a display-only convenience: such upstreams canonicalize
-  // the slash form back to "__" on each request, so requests still round-trip.
-  // Defaults false; the wizard turns it on only for the llmproxy template.
-  rewriteDoubleUnderscore?: boolean;
 }
 
 interface ExtensionConfig {
@@ -104,13 +97,6 @@ const TEMPLATES: Record<string, {
   modelsIdField?: string;
   /** Keep only models whose task.name matches this string (case-insensitive). */
   modelsKeepTask?: string;
-  /**
-   * Rewrite the first "__" of each model id to "/" before registering with pi,
-   * for display compatibility — ids from "provider__model" upstreams (e.g.
-   * llmproxy) then show in the familiar "provider/model" slash form. Display-only;
-   * default false; set true where you want the slash-form presentation.
-   */
-  rewriteDoubleUnderscore?: boolean;
 }> = {
   openrouter: {
     displayName: "OpenRouter",
@@ -302,10 +288,6 @@ const TEMPLATES: Record<string, {
     baseUrl: "http://localhost:8080/v1",
     keyless: true,
     promptUrl: true,
-    // llmproxy advertises "provider__model" ids (and virtuals like
-    // "llmproxy__free"). Rewrite the first "__" to "/" here so they display in
-    // the familiar slash form; llmproxy canonicalizes it back on requests.
-    rewriteDoubleUnderscore: true,
   },
   vercel: {
     displayName: "Vercel AI Gateway",
@@ -553,21 +535,9 @@ async function fetchModels(
 // Provider registration helpers
 // ─────────────────────────────────────────────────────────────────────────────
 
-/**
- * Rewrite the FIRST "__" of a model id to "/". Only the first occurrence is
- * touched: the provider/virtual segment that precedes it never contains "__" or
- * "/", so this is the exact inverse of the slash→"__" canonicalization llmproxy
- * applies on the request side, keeping ids round-trippable. Ids without "__" are
- * returned unchanged.
- */
-function rewriteFirstDoubleUnderscore(id: string): string {
-  const i = id.indexOf("__");
-  return i === -1 ? id : `${id.slice(0, i)}/${id.slice(i + 2)}`;
-}
-
-function buildProviderModels(models: CachedModel[], rewriteDoubleUnderscore = false) {
+function buildProviderModels(models: CachedModel[]) {
   return models.map((m) => {
-    const id = rewriteDoubleUnderscore ? rewriteFirstDoubleUnderscore(m.id) : m.id;
+    const id = m.id;
     return {
       id,
       name: id,
@@ -585,16 +555,12 @@ function compatKey(key: string): string {
 }
 
 function registerProvider(pi: ExtensionAPI, key: string, p: ProviderConfig): void {
-  // Honor the persisted per-provider flag; fall back to the template default for
-  // this key so providers logged in before the flag existed (e.g. an existing
-  // llmproxy provider) still get the rewrite without a re-login.
-  const rewrite = p.rewriteDoubleUnderscore ?? TEMPLATES[key]?.rewriteDoubleUnderscore ?? false;
   pi.registerProvider(compatKey(key), {
     name: `compat/${key.replace(/_/g, "-")}`,
     baseUrl: p.baseUrl,
     apiKey: p.apiKey ?? (isLocalUrl(p.baseUrl) ? "local" : ""),
     api: "openai-completions" as const,
-    models: buildProviderModels(p.cachedModels, rewrite),
+    models: buildProviderModels(p.cachedModels),
   });
 }
 
@@ -837,7 +803,6 @@ export default async function (pi: ExtensionAPI) {
         modelsUrl,
         modelsIdField: tpl.modelsIdField,
         modelsKeepTask: tpl.modelsKeepTask,
-        rewriteDoubleUnderscore: tpl.rewriteDoubleUnderscore,
       };
       saveConfig(config);
       registerProvider(pi, key, config.providers[key]);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@billjr99/pi-openai-compat",
-  "version": "1.1.17",
+  "version": "1.1.18",
   "description": "pi-coding-agent extension: OpenAI-compatible endpoint support (OpenRouter, NVIDIA NIM, Nous Portal, Ollama, custom)",
   "author": "Bill Mongan <https://github.com/BillJr99>",
   "license": "MIT",