@goodandready/dsh-opencode-provider 0.5.27 → 0.5.29

package/CHANGELOG.md

@@ -1,5 +1,17 @@
 # Changelog
 
+## 0.5.29 — release 6: catalog refresh and Free-lane decision
+
+- **#77 — catalog refresh**: `catalogRefreshMinutes` (1–10,080, default 10) controls the live `/models` cache; explicit `/models/refresh` still bypasses it. Exposed in Settings.
+- **#76 — Free-lane decision**: re-probed keyless `big-pickle` on 2026-11-05; HTTP 429 `FreeUsageLimitError`. No keyless route or header-spoofing lane is added; the keyed Console/Zen path remains supported. Evidence recorded in DESIGN/README.
+- 137 tests.
+
+## 0.5.28 — release 5: image policy and preparation timeout
+
+- **#73 — image policy**: `maxImages`, `maxRequestImageBytes`, `requestImagePixelBudget` and `requestImageMaxBytes` are configurable in Settings; defaults preserve the previous 50 images / 24 MiB / 1568px / 1 MiB behavior. Count and pixel policies are covered by tests.
+- **#78 — preparation timeout**: `requestPreparationTimeoutMs` (default 60s) bounds image resolution before dispatch, independently of the response-header timeout; caller aborts still win, and timeouts surface as `TIMEOUT`.
+- 136 tests.
+
 ## 0.5.27 — release 4: catalog freshness
 
 - **#86 — new-model badge**: the live catalog now keeps `release_date` (or `created`) metadata; Settings marks models released within the last seven UTC days with a NEW badge and date tooltip. Dates are dropped safely when the gateway omits or malforms them.

package/README.md

@@ -132,12 +132,12 @@ npm pack
 Install the **packed tarball** into a profile:
 
 ```bash
-dsh plugin --profile web add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.27.tgz
+dsh plugin --profile web add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.29.tgz
 # DSH Desktop:
-dsh plugin --profile desktop add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.27.tgz
+dsh plugin --profile desktop add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.29.tgz
 
 # Or install the published package:
-dsh plugin --profile web add @goodandready/dsh-opencode-provider@0.5.27
+dsh plugin --profile web add @goodandready/dsh-opencode-provider@0.5.29
 ```
 
 Restart dsh (or DSH Desktop), then pick **OpenCode (plugin)** or **OpenCode Go (plugin)** in the model picker.
@@ -310,7 +310,7 @@ GitHub Actions runs `npm run check` and `npm pack` on push/PR (`.github/workflow
 - Subscription request counts are OpenCode's published estimates, not enforced quotas. They can change when OpenCode updates the Go pricing table; the gateway is the final authority.
 - Usage-by-model aggregates the local projection cache and attributes a whole session to its last-used model; a session that switched models is therefore approximate. It never reads prompt text or titles.
 - `opencode-direct` (Console) billing shows no per-model allowance; the Go route is the primary target.
-- The anonymous free lane (`Bearer public` + CLI impersonation) is not implemented.
+- The anonymous free lane (`Bearer public` + CLI impersonation) is not implemented. A keyless `big-pickle` probe on 2026-11-05 returned HTTP 429 `FreeUsageLimitError`; use the keyed Console/Zen route for free models.
 - The UI requires a web profile (the Host needs `ctx.webServer`); the TUI keeps the `/opencode` command and the model routes.
 - Write routes are loopback-only; a web profile exposed to a LAN can read state but cannot save configuration or keys until the host auth token is integrated.
 

package/README.ru.md

@@ -131,12 +131,12 @@ npm pack
 Установка из **упакованного tarball**:
 
 ```bash
-dsh plugin --profile web add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.27.tgz
+dsh plugin --profile web add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.29.tgz
 # DSH Desktop:
-dsh plugin --profile desktop add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.27.tgz
+dsh plugin --profile desktop add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.29.tgz
 
 # Либо установка опубликованного пакета из npm:
-dsh plugin --profile web add @goodandready/dsh-opencode-provider@0.5.27
+dsh plugin --profile web add @goodandready/dsh-opencode-provider@0.5.29
 ```
 
 После установки перезапустите dsh (или DSH Desktop) и выберите **OpenCode (plugin)** или **OpenCode Go (plugin)** в списке моделей.

package/README.zh.md

@@ -126,12 +126,12 @@ npm pack
 通过**打包文件 (.tgz)**安装：
 
 ```bash
-dsh plugin --profile web add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.27.tgz
+dsh plugin --profile web add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.29.tgz
 # DSH 桌面端：
-dsh plugin --profile desktop add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.27.tgz
+dsh plugin --profile desktop add /absolute/path/to/goodandready-dsh-opencode-provider-0.5.29.tgz
 
 # 或从 npm 安装已发布版本：
-dsh plugin --profile web add @goodandready/dsh-opencode-provider@0.5.27
+dsh plugin --profile web add @goodandready/dsh-opencode-provider@0.5.29
 ```
 
 安装完成后重启 dsh 或 DSH 桌面端，在模型选择器中选择 **OpenCode (plugin)** 或 **OpenCode Go (plugin)** 即可。

package/lib/client.js

@@ -181,6 +181,12 @@ window.__ModuleLoader__.load({
 				usageUrl: "Usage endpoint",
 				requestTimeoutMs: "Header timeout (ms)",
 				streamIdleTimeoutMs: "Stream idle timeout (ms)",
+				requestPreparationTimeoutMs: "Preparation timeout (ms)",
+				maxImages: "Max images per request",
+				maxRequestImageBytes: "Request image budget (bytes)",
+				requestImagePixelBudget: "Image pixel budget",
+				requestImageMaxBytes: "Per-image max bytes",
+				catalogRefreshMinutes: "Catalog refresh (minutes)",
 				proxyUrl: "Proxy URL",
 				proxyUrlHint: "HTTP, HTTPS or SOCKS5 proxy for inference, catalog and usage. Empty disables it. Credentials are not supported.",
 				maxOutputTokens: "Output limit (empty = model default, 0 = omit)",
@@ -409,6 +415,12 @@ window.__ModuleLoader__.load({
 				usageUrl: "用量端点",
 				requestTimeoutMs: "请求头超时（毫秒）",
 				streamIdleTimeoutMs: "流空闲超时（毫秒）",
+				requestPreparationTimeoutMs: "准备超时（毫秒）",
+				maxImages: "每请求最大图片数",
+				maxRequestImageBytes: "请求图片预算（字节）",
+				requestImagePixelBudget: "图片像素预算",
+				requestImageMaxBytes: "单图最大字节数",
+				catalogRefreshMinutes: "目录刷新间隔（分钟）",
 				proxyUrl: "代理地址",
 				proxyUrlHint: "用于推理、目录和用量的 HTTP、HTTPS 或 SOCKS5 代理。留空则禁用。不支持代理认证。",
 				maxOutputTokens: "输出上限（空 = 模型默认，0 = 省略）",
@@ -1696,7 +1708,13 @@ window.__ModuleLoader__.load({
 				openCodeAuthEntry: state.config.openCodeAuthEntry,
 				requestTimeoutMs: String(state.config.requestTimeoutMs),
 				streamIdleTimeoutMs: String(state.config.streamIdleTimeoutMs),
+				requestPreparationTimeoutMs: String(state.config.requestPreparationTimeoutMs),
+				maxImages: String(state.config.maxImages),
+				maxRequestImageBytes: String(state.config.maxRequestImageBytes),
+				requestImagePixelBudget: String(state.config.requestImagePixelBudget),
+				requestImageMaxBytes: String(state.config.requestImageMaxBytes),
 				proxyUrl: state.config.proxyUrl,
+				catalogRefreshMinutes: String(state.config.catalogRefreshMinutes),
 				maxOutputTokens: state.config.maxOutputTokens === null ? "" : String(state.config.maxOutputTokens),
 				contextWindow: state.config.contextWindow === null ? "" : String(state.config.contextWindow),
 				contextWindowOverrides: Object.fromEntries(Object.entries(state.config.contextWindowOverrides).map(([id, value]) => [id, String(value)])),
@@ -1743,7 +1761,13 @@ window.__ModuleLoader__.load({
 				openCodeAuthEntry: draft.openCodeAuthEntry.trim(),
 				requestTimeoutMs: numberOrNull(draft.requestTimeoutMs) ?? 3e5,
 				streamIdleTimeoutMs: numberOrNull(draft.streamIdleTimeoutMs) ?? 3e5,
+				requestPreparationTimeoutMs: numberOrNull(draft.requestPreparationTimeoutMs) ?? 6e4,
+				maxImages: numberOrNull(draft.maxImages) ?? 50,
+				maxRequestImageBytes: numberOrNull(draft.maxRequestImageBytes) ?? 25165824,
+				requestImagePixelBudget: numberOrNull(draft.requestImagePixelBudget) ?? 2458624,
+				requestImageMaxBytes: numberOrNull(draft.requestImageMaxBytes) ?? 1048576,
 				proxyUrl: draft.proxyUrl.trim(),
+				catalogRefreshMinutes: numberOrNull(draft.catalogRefreshMinutes) ?? 10,
 				maxOutputTokens: numberOrNull(draft.maxOutputTokens),
 				contextWindow: numberOrNull(draft.contextWindow),
 				contextWindowOverrides: numberRecordOrNull(draft.contextWindowOverrides),
@@ -3148,6 +3172,42 @@ window.__ModuleLoader__.load({
 									value: draft.streamIdleTimeoutMs,
 									onChange: (value) => update({ streamIdleTimeoutMs: value })
 								}),
+								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
+									label: t("requestPreparationTimeoutMs"),
+									type: "number",
+									value: draft.requestPreparationTimeoutMs,
+									onChange: (value) => update({ requestPreparationTimeoutMs: value })
+								}),
+								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
+									label: t("maxImages"),
+									type: "number",
+									value: draft.maxImages,
+									onChange: (value) => update({ maxImages: value })
+								}),
+								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
+									label: t("maxRequestImageBytes"),
+									type: "number",
+									value: draft.maxRequestImageBytes,
+									onChange: (value) => update({ maxRequestImageBytes: value })
+								}),
+								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
+									label: t("requestImagePixelBudget"),
+									type: "number",
+									value: draft.requestImagePixelBudget,
+									onChange: (value) => update({ requestImagePixelBudget: value })
+								}),
+								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
+									label: t("requestImageMaxBytes"),
+									type: "number",
+									value: draft.requestImageMaxBytes,
+									onChange: (value) => update({ requestImageMaxBytes: value })
+								}),
+								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
+									label: t("catalogRefreshMinutes"),
+									type: "number",
+									value: draft.catalogRefreshMinutes,
+									onChange: (value) => update({ catalogRefreshMinutes: value })
+								}),
 								/* @__PURE__ */ (0, react_jsx_runtime.jsx)(Field, {
 									label: t("proxyUrl"),
 									hint: t("proxyUrlHint"),

package/lib/index.d.ts

@@ -37,8 +37,20 @@ interface Config {
   requestTimeoutMs?: number;
   /** Milliseconds a stream may stall; defaults to 300s. */
   streamIdleTimeoutMs?: number;
+  /** Milliseconds allowed for image resolution and request preparation. */
+  requestPreparationTimeoutMs?: number;
+  /** Request-wide image count cap; defaults to 50. */
+  maxImages?: number;
+  /** Request-wide base64 image budget; defaults to 24 MiB. */
+  maxRequestImageBytes?: number;
+  /** Total-pixel budget for one normalized image. */
+  requestImagePixelBudget?: number;
+  /** Raw encoded-byte target for one image before base64 expansion. */
+  requestImageMaxBytes?: number;
   /** HTTP/HTTPS/SOCKS5 proxy for inference, catalog and usage requests. */
   proxyUrl?: string;
+  /** Live catalog cache lifetime in minutes; defaults to 10. */
+  catalogRefreshMinutes?: number;
   /** Visible-model allowlist; an empty list means "show everything". */
   visibleModels?: string[];
   /** Per-model visibility overrides; wins over `visibleModels`. */
@@ -85,7 +97,13 @@ interface ResolvedConfig {
   openCodeAuthEntry: string;
   requestTimeoutMs: number;
   streamIdleTimeoutMs: number;
+  requestPreparationTimeoutMs: number;
+  maxImages: number;
+  maxRequestImageBytes: number;
+  requestImagePixelBudget: number;
+  requestImageMaxBytes: number;
   proxyUrl: string;
+  catalogRefreshMinutes: number;
   visibleModels: readonly string[];
   modelVisibility: Readonly<Record<string, boolean>>;
   filterModelsByPlan: boolean;

package/lib/index.js

@@ -81,10 +81,14 @@ const DEFAULT_GO_API_BASE = "https://opencode.ai/inference/go";
 const DEFAULT_GO_ZEN_BASE = "https://opencode.ai/zen/go/v1";
 const DEFAULT_USAGE_URL = "https://opencode.ai/zen/go/v1/usage";
 const DEFAULT_AUTH_ENTRY = "opencode-go";
-const DEFAULT_USER_AGENT = "dsh-opencode-provider/0.5.27";
+const DEFAULT_USER_AGENT = "dsh-opencode-provider/0.5.29";
 const DEFAULT_CONTEXT_WINDOW = 262144;
 const DEFAULT_MAX_TOKENS = 32768;
 const DEFAULT_GENERATE_MAX_TOKENS = 65536;
+const DEFAULT_MAX_REQUEST_IMAGE_BYTES = 25165824;
+const DEFAULT_REQUEST_IMAGE_PIXEL_BUDGET = 2458624;
+const DEFAULT_REQUEST_IMAGE_MAX_BYTES = 1048576;
+const DEFAULT_REQUEST_PREPARATION_TIMEOUT_MS = 6e4;
 function markVolatile(schema) {
 	const mark = schema.volatile;
 	return typeof mark === "function" ? mark.call(schema) : schema;
@@ -99,7 +103,13 @@ const Config = z.object({
 	openCodeAuthEntry: markVolatile(z.string().default(DEFAULT_AUTH_ENTRY)),
 	requestTimeoutMs: markVolatile(z.number().min(1).max(2147483647).default(3e5)),
 	streamIdleTimeoutMs: markVolatile(z.number().min(1).max(2147483647).default(3e5)),
+	requestPreparationTimeoutMs: markVolatile(z.number().min(1).max(2147483647).default(DEFAULT_REQUEST_PREPARATION_TIMEOUT_MS)),
+	maxImages: markVolatile(z.number().min(1).max(1e4).step(1).default(50)),
+	maxRequestImageBytes: markVolatile(z.number().min(1).max(1e9).step(1).default(DEFAULT_MAX_REQUEST_IMAGE_BYTES)),
+	requestImagePixelBudget: markVolatile(z.number().min(1).max(1e8).step(1).default(DEFAULT_REQUEST_IMAGE_PIXEL_BUDGET)),
+	requestImageMaxBytes: markVolatile(z.number().min(1).max(1e8).step(1).default(DEFAULT_REQUEST_IMAGE_MAX_BYTES)),
 	proxyUrl: markVolatile(z.string().default("")),
+	catalogRefreshMinutes: markVolatile(z.number().min(1).max(10080).step(1).default(10)),
 	visibleModels: markVolatile(z.array(z.string()).default([])),
 	modelVisibility: markVolatile(z.dict(z.boolean()).default({})),
 	filterModelsByPlan: markVolatile(z.boolean().default(true)),
@@ -212,7 +222,13 @@ function resolveConfig(config) {
 		openCodeAuthEntry: typeof config.openCodeAuthEntry === "string" && config.openCodeAuthEntry !== "" ? config.openCodeAuthEntry : DEFAULT_AUTH_ENTRY,
 		requestTimeoutMs: positiveInt(config.requestTimeoutMs, 3e5),
 		streamIdleTimeoutMs: positiveInt(config.streamIdleTimeoutMs, 3e5),
+		requestPreparationTimeoutMs: positiveInt(config.requestPreparationTimeoutMs, DEFAULT_REQUEST_PREPARATION_TIMEOUT_MS),
+		maxImages: positiveInt(config.maxImages, 50),
+		maxRequestImageBytes: positiveInt(config.maxRequestImageBytes, DEFAULT_MAX_REQUEST_IMAGE_BYTES),
+		requestImagePixelBudget: positiveInt(config.requestImagePixelBudget, DEFAULT_REQUEST_IMAGE_PIXEL_BUDGET),
+		requestImageMaxBytes: positiveInt(config.requestImageMaxBytes, DEFAULT_REQUEST_IMAGE_MAX_BYTES),
 		proxyUrl: normalizeProxyUrl(config.proxyUrl),
+		catalogRefreshMinutes: positiveInt(config.catalogRefreshMinutes, 10),
 		visibleModels: Array.isArray(config.visibleModels) ? config.visibleModels.filter((id) => typeof id === "string" && id !== "") : [],
 		modelVisibility: booleanRecord(config.modelVisibility),
 		filterModelsByPlan: config.filterModelsByPlan !== false,
@@ -1052,7 +1068,6 @@ function describeLimitsShort(limits) {
 	return `Go ${formatRequests(limits.go.fiveHour)}/5h · ${formatRequests(limits.go.weekly)}/wk`;
 }
 // src/catalog.ts
-const CATALOG_TTL_MS = 6e5;
 const CATALOG_TIMEOUT_MS = 15e3;
 const catalogCache = /* @__PURE__ */ new Map();
 const FALLBACK_META = {
@@ -1130,7 +1145,7 @@ async function fetchLiveIds(provider, config, apiKey, fetchImpl) {
 	const url = modelsUrlFor(provider, config);
 	const cacheKey = `${provider}|${url}`;
 	const cached = catalogCache.get(cacheKey);
-	if (cached !== void 0 && Date.now() - cached.at < CATALOG_TTL_MS) return cached.ids;
+	if (cached !== void 0 && Date.now() - cached.at < config.catalogRefreshMinutes * 6e4) return cached.ids;
 	const headers = {
 		accept: "application/json",
 		"user-agent": config.userAgent,
@@ -1292,16 +1307,21 @@ function classifyStreamError(value, status = 200) {
 }
 // src/images.ts
 const MAX_LONG_EDGE = 1568;
-const MAX_BYTES = 1048576;
-/** Total base64 image budget for one request; leaves room under the gateway cap. */
-const MAX_REQUEST_IMAGE_BYTES = 25165824;
-const MAX_REQUEST_IMAGES = 50;
+function limitsOf(policy) {
+	return {
+		maxImages: policy?.maxImages ?? 50,
+		maxRequestImageBytes: policy?.maxRequestImageBytes ?? 25165824,
+		requestImagePixelBudget: policy?.requestImagePixelBudget ?? 2458624,
+		requestImageMaxBytes: policy?.requestImageMaxBytes ?? 1048576
+	};
+}
 /**
 * Keep the newest images under one request-wide budget. The oldest images are
 * dropped first; the transports render them as an "image omitted" placeholder
 * instead of failing the whole request.
 */
-function selectRequestImages(refs, occurrences) {
+function selectRequestImages(refs, occurrences, policy) {
+	const limits = limitsOf(policy);
 	const unique = [];
 	const seen = /* @__PURE__ */ new Set();
 	for (const ref of refs) {
@@ -1314,7 +1334,7 @@ function selectRequestImages(refs, occurrences) {
 	let start = 0;
 	let totalImages = unique.reduce((sum, ref) => sum + countOf(ref), 0);
 	let bytes = unique.reduce((sum, ref) => sum + encoded(ref.bytes) * countOf(ref), 0);
-	while (start < unique.length && (totalImages > MAX_REQUEST_IMAGES || bytes > MAX_REQUEST_IMAGE_BYTES)) {
+	while (start < unique.length && (totalImages > limits.maxImages || bytes > limits.maxRequestImageBytes)) {
 		const dropped = unique[start];
 		if (dropped !== void 0) {
 			const count = countOf(dropped);
@@ -1333,20 +1353,23 @@ function selectRequestImages(refs, occurrences) {
 		dropped: unique.slice(0, start)
 	};
 }
-function targetOf(ref) {
+function targetOf(ref, policy) {
+	const limits = limitsOf(policy);
 	const longEdge = Math.max(ref.width, ref.height);
-	const scale = longEdge > MAX_LONG_EDGE ? MAX_LONG_EDGE / longEdge : 1;
+	const pixelEdge = Math.max(1, Math.floor(Math.sqrt(limits.requestImagePixelBudget)));
+	const maxEdge = Math.min(MAX_LONG_EDGE, pixelEdge);
+	const scale = longEdge > maxEdge ? maxEdge / longEdge : 1;
 	return {
 		width: Math.max(1, Math.round(ref.width * scale)),
 		height: Math.max(1, Math.round(ref.height * scale)),
-		maxBytes: MAX_BYTES
+		maxBytes: limits.requestImageMaxBytes
 	};
 }
 function hasImages(options) {
 	return options.messages.some((message) => message.content.some((block) => block.type === "image" && block.offloaded !== true));
 }
 /** Decode the distinct request images that fit the request-wide budget. */
-async function resolveImages(options, store, signal) {
+async function resolveImages(options, store, signal, policy) {
 	const refs = [];
 	const counts = /* @__PURE__ */ new Map();
 	const seen = /* @__PURE__ */ new Set();
@@ -1359,8 +1382,8 @@ async function resolveImages(options, store, signal) {
 		refs.push(ref);
 	}
 	const images = /* @__PURE__ */ new Map();
-	for (const ref of selectRequestImages(refs, counts).kept) {
-		const request = await store.readImageRequest(ref, targetOf(ref), signal);
+	for (const ref of selectRequestImages(refs, counts, policy).kept) {
+		const request = await store.readImageRequest(ref, targetOf(ref, policy), signal);
 		const base64 = Buffer.from(request.data).toString("base64");
 		images.set(ref.attachmentId, {
 			mediaType: request.mediaType,
@@ -3369,7 +3392,25 @@ var OpenCodeAdapter = class extends LlmAdapter {
 		if (hasImages(options)) {
 			const store = this.deps.resolveAttachments?.();
 			if (store === void 0) throw new LlmError("llm-opencode: image input requires the durable attachment service", "UNSUPPORTED_CONTENT");
-			images = await resolveImages(options, store, options.signal);
+			const preparation = new AbortController();
+			let timedOut = false;
+			const onAbort = () => preparation.abort(options.signal?.reason);
+			options.signal?.addEventListener("abort", onAbort, { once: true });
+			const timer = setTimeout(() => {
+				timedOut = true;
+				preparation.abort();
+			}, config.requestPreparationTimeoutMs);
+			timer.unref?.();
+			try {
+				images = await resolveImages(options, store, preparation.signal, config);
+			} catch (error) {
+				if (options.signal?.aborted === true) throw options.signal.reason ?? error;
+				if (timedOut) throw new LlmError(`OpenCode image preparation exceeded ${config.requestPreparationTimeoutMs}ms`, "TIMEOUT", { cause: error });
+				throw error;
+			} finally {
+				clearTimeout(timer);
+				options.signal?.removeEventListener("abort", onAbort);
+			}
 		}
 		let bodyOptions = options;
 		if (options.tools !== void 0) {
@@ -4748,7 +4789,13 @@ function publicConfig(config) {
 		openCodeAuthEntry: config.openCodeAuthEntry,
 		requestTimeoutMs: config.requestTimeoutMs,
 		streamIdleTimeoutMs: config.streamIdleTimeoutMs,
+		requestPreparationTimeoutMs: config.requestPreparationTimeoutMs,
+		maxImages: config.maxImages,
+		maxRequestImageBytes: config.maxRequestImageBytes,
+		requestImagePixelBudget: config.requestImagePixelBudget,
+		requestImageMaxBytes: config.requestImageMaxBytes,
 		proxyUrl: config.proxyUrl,
+		catalogRefreshMinutes: config.catalogRefreshMinutes,
 		visibleModels: config.visibleModels,
 		modelVisibility: config.modelVisibility,
 		filterModelsByPlan: config.filterModelsByPlan,
@@ -4944,7 +4991,13 @@ function configOps(patch) {
 	if ("openCodeAuthEntry" in patch) set("openCodeAuthEntry", patch.openCodeAuthEntry);
 	if ("requestTimeoutMs" in patch) set("requestTimeoutMs", patch.requestTimeoutMs);
 	if ("streamIdleTimeoutMs" in patch) set("streamIdleTimeoutMs", patch.streamIdleTimeoutMs);
+	if ("requestPreparationTimeoutMs" in patch) set("requestPreparationTimeoutMs", patch.requestPreparationTimeoutMs);
+	if ("maxImages" in patch) set("maxImages", patch.maxImages);
+	if ("maxRequestImageBytes" in patch) set("maxRequestImageBytes", patch.maxRequestImageBytes);
+	if ("requestImagePixelBudget" in patch) set("requestImagePixelBudget", patch.requestImagePixelBudget);
+	if ("requestImageMaxBytes" in patch) set("requestImageMaxBytes", patch.requestImageMaxBytes);
 	if ("proxyUrl" in patch) set("proxyUrl", normalizeProxyUrl(typeof patch.proxyUrl === "string" ? patch.proxyUrl : ""));
+	if ("catalogRefreshMinutes" in patch) set("catalogRefreshMinutes", patch.catalogRefreshMinutes);
 	if ("visibleModels" in patch) set("visibleModels", cleanStringArray(patch.visibleModels));
 	if ("modelVisibility" in patch) set("modelVisibility", cleanBooleanRecord(patch.modelVisibility));
 	if ("filterModelsByPlan" in patch) set("filterModelsByPlan", patch.filterModelsByPlan === true);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@goodandready/dsh-opencode-provider",
-  "version": "0.5.27",
+  "version": "0.5.29",
   "description": "OpenCode / OpenCode Go LLM provider plugin for DeepSeek Harness (dsh).",
   "license": "MIT",
   "type": "module",