fix(providers): preserve explicit vision restrictions

src/providers/model-rename-startup.ts

@@ -1,26 +1,24 @@
 import { mutatePersistedConfig } from "../config";
 import { projectModelRenames } from "./model-rename-migration";
 import { projectStaleContextWindows } from "./stale-context-window-migration";
-import { projectStaleVisionClassifications } from "./stale-vision-classification-migration";
 import { projectDevinCliAuthMode } from "./devin-cli-authmode-migration";
 import type { OcxConfig } from "../types";
 
 /**
- * The startup projection: registry model renames, then the context-window and
- * vision-classification repairs. All three fix a saved row the registry can no longer reach on its own —
+ * The startup projection: registry model renames, then context-window and auth-mode repairs.
+ * These fix a saved row the registry can no longer reach on its own —
  * `enrichProviderFromRegistry` backfills a missing field and never rewrites a
  * present one — so they share this pass rather than adding a second boot step
  * with its own persistence, adopt, and failure handling.
  */
 export function projectStartupConfigRepairs(config: OcxConfig): ReturnType<typeof projectModelRenames> {
   const renames = projectModelRenames(config);
   const windows = projectStaleContextWindows(renames.config);
-  const vision = projectStaleVisionClassifications(windows.config);
-  const devinCli = projectDevinCliAuthMode(vision.config);
+  const devinCli = projectDevinCliAuthMode(windows.config);
   return {
     config: devinCli.config,
-    changed: renames.changed || windows.changed || vision.changed || devinCli.changed,
-    warnings: [...renames.warnings, ...windows.warnings, ...vision.warnings, ...devinCli.warnings],
+    changed: renames.changed || windows.changed || devinCli.changed,
+    warnings: [...renames.warnings, ...windows.warnings, ...devinCli.warnings],
   };
 }
 

src/providers/registry/entries-core.ts

@@ -718,12 +718,9 @@ export const PROVIDER_REGISTRY_CORE: readonly ProviderRegistryEntry[] = [
       // HTTP 400 "Model only supports text input", which is what keeps the two
       // distinct here rather than collapsing them.
       //
-      // The declaration is what reaches an EXISTING install: derive.ts fills
-      // noVisionModels all-or-nothing, so a config persisted while the stale list
-      // was current keeps it forever, and modelInputModalities is filled per-key
-      // BENEATH the saved value. Both halves are repaired by
-      // stale-vision-classification-migration.ts; correcting the registry alone
-      // would fix new installs and leave existing ones stripping images.
+      // derive.ts fills missing registry metadata but does not replace saved values.
+      // That makes this declaration the default for new rows while preserving an
+      // existing install's operator-editable image-routing restrictions.
       "deepseek-v4.1-flash": ["text", "image"],
       // Muse Spark Contributor is natively multimodal on Zen Go: it accepts input_image
       // parts over /responses (probed 2026-08-26). Without this declaration the catalog
@@ -1258,4 +1255,3 @@ export const PROVIDER_REGISTRY_CORE: readonly ProviderRegistryEntry[] = [
     note: "Serverless Inference subscription API. Live discovery exposes only kimi-k2-instruct because Vultr documents it as the sole tool-calling model.",
   },
 ];
-

structure/providers-and-adapters.md

@@ -78,27 +78,13 @@ answers HTTP 400 "Model only supports text input" and stays sidecar-backed. The
 (`opencode-zen`, `opencode-free`) were not measurable (HTTP 402) and keep their existing
 classification — an unverified tier is not evidence.
 
-Because `enrichProviderFromRegistry` fills `noVisionModels` all-or-nothing and fills
-`modelInputModalities` per-key beneath the saved value, both halves of a stale classification are
-frozen into any config saved while it was current. `src/providers/stale-vision-classification-migration.ts`
-repairs exactly those two saved values and runs inside the shared startup repair pass in
-`src/providers/model-rename-startup.ts`. Correcting the registry alone fixes new installs only.
-
-It covers both states that reach a running process, because the sidecar predicate reads
-`noVisionModels` before `modelInputModalities`: the full stale pair (modalities still the stale
-declaration and the id listed, both rewritten) and the half-repaired row (modalities already
-corrected but the id still listed, where removing the name is what stops the image from being
-stripped). The paired modality declaration is the guard in both cases, which is why a name listed
-without one is left alone — that row is either a half-finished repair or a deliberate operator
-entry, and the projection does not guess which. The row must also still be the registry's own:
-identity resolves through `providerMatchesRegistryTransport`, the rule `enrichProviderFromRegistry`
-applies before it writes registry metadata, plus the entry's adapter. `opencode-go` is a pinned
-key preset without `preserveCustomDestination`, so its id alone claims a row — exactly as it does
-for enrichment — and an entry that opts into destination preservation narrows the projection with
-it. `modelCapabilities` is never written: it is the
-axis that outranks every source here, so it is where a deliberate text-only override belongs
-(`ocx provider edit <provider> --model <id> --text-only` writes it) and the one declaration a
-restart cannot take back.
+Because `enrichProviderFromRegistry` is fill-only, the corrected classification applies to new
+rows and missing values. Startup does not rewrite an existing `noVisionModels` entry or
+`modelInputModalities` value: those fields are operator-editable request-routing policy, and a
+saved registry seed is indistinguishable from an intentional restriction without provenance.
+Existing users can opt into the corrected native-vision classification by removing those saved
+overrides. `modelCapabilities` remains the highest-precedence per-model axis, and
+`ocx provider edit <provider> --model <id> --text-only` writes an explicit restriction there.
 
 The BigModel Coding Plan Responses preset uses the separately documented
 `https://open.bigmodel.cn/api/v1` transport and a static catalog. Its provider row