copilot-relay 0.1.0 → 0.2.0

package/README.md

@@ -37,7 +37,7 @@ Routing:
 | Requested model | Upstream model |
 | --- | --- |
 | contains `opus` | `claude-opus-4.8` |
-| anything else | `gpt-5.5` |
+| anything else | `gpt-5.6-sol` |
 
 ## Install & run
 
@@ -61,7 +61,7 @@ copilotBaseUrl: https://api.githubcopilot.com
 claudeSetup: true
 logLevel: info
 logRetentionDays: 3
-thinkEffort: xhigh
+thinkEffort: max
 upstreamTimeoutSeconds: 180
 webSearchBackend:
 ```
@@ -76,7 +76,7 @@ webSearchBackend:
 
 Any other `logLevel` value is invalid and stops startup.
 
-Valid `thinkEffort`: `none`, `low`, `medium`, `high`, `xhigh`.
+Valid `thinkEffort`: `none`, `low`, `medium`, `high`, `xhigh`, `max`.
 
 `upstreamTimeoutSeconds` controls the maximum time a single Claude request can
 spend waiting on upstream Copilot calls, including chat, Responses, preflight,

package/config.default.yaml

@@ -32,8 +32,8 @@ logLevel: info
 logRetentionDays: 3
 
 # Default upstream thinking/reasoning effort sent to Copilot.
-# Valid values: none, low, medium, high, xhigh.
-thinkEffort: xhigh
+# Valid values: none, low, medium, high, xhigh, max.
+thinkEffort: max
 
 # Max seconds to wait for one Claude request's upstream Copilot calls.
 upstreamTimeoutSeconds: 180
@@ -45,7 +45,7 @@ webSearchBackend:
 # Model routing: requests containing "opus" use opusModel; all others use gptModel.
 #
 # Upstream Copilot model used for non-Opus requests.
-gptModel: gpt-5.5
+gptModel: gpt-5.6-sol
 
 # Upstream Copilot model used for requests containing "opus".
 opusModel: claude-opus-4.8

package/dist/main.js

@@ -343,12 +343,12 @@ const runtimeState = {};
 
 //#endregion
 //#region src/lib/models.ts
-const defaultReasoningEffort = "xhigh";
+const defaultReasoningEffort = "max";
 const defaultModelRouting = {
-	gptModel: "gpt-5.5",
+	gptModel: "gpt-5.6-sol",
 	opusModel: "claude-opus-4.8"
 };
-const isReasoningEffort = (value) => value === "none" || value === "low" || value === "medium" || value === "high" || value === "xhigh";
+const isReasoningEffort = (value) => value === "none" || value === "low" || value === "medium" || value === "high" || value === "xhigh" || value === "max";
 const getModelRouting = () => runtimeState.modelRouting ?? defaultModelRouting;
 const getExposedModelIds = () => {
 	const routing = getModelRouting();
@@ -369,7 +369,7 @@ const logLevels = [
 const defaultConfig = {
 	claudeSetup: true,
 	copilotBaseUrl: "https://api.githubcopilot.com",
-	gptModel: "gpt-5.5",
+	gptModel: "gpt-5.6-sol",
 	host: "127.0.0.1",
 	logLevel: "info",
 	logRetentionDays: 3,
@@ -534,7 +534,7 @@ const serializeConfig = (config) => [
 	"# Number of days to keep files in ~/.copilot-relay/logs.",
 	`logRetentionDays: ${config.logRetentionDays}`,
 	"",
-	"# Default upstream thinking/reasoning effort: none, low, medium, high, xhigh.",
+	"# Default upstream thinking/reasoning effort: none, low, medium, high, xhigh, max.",
 	`thinkEffort: ${config.thinkEffort}`,
 	"",
 	"# Max seconds to wait for a single Claude request's upstream Copilot calls.",
@@ -982,7 +982,7 @@ const fetchCopilot = async (provider, path$1, init, options = {}) => {
 
 //#endregion
 //#region src/copilot/responses.ts
-const responsesOnlyModelPattern = /^(?:gpt-5\.5)(?:-|$)/i;
+const responsesOnlyModelPattern = /^(?:gpt-5\.5|gpt-5\.6)(?:-|$)/i;
 function shouldUseResponsesApiForModel(model) {
 	return responsesOnlyModelPattern.test(model);
 }
@@ -2477,7 +2477,6 @@ const getTokenCount = async (payload, model) => {
 //#endregion
 //#region src/routes/claude.ts
 const claudeRoutes = new Hono();
-const streamKeepAliveIntervalMs = 15e3;
 const createTokenCountModel = (modelId) => ({
 	capabilities: { tokenizer: "o200k_base" },
 	id: modelId
@@ -2555,17 +2554,6 @@ const createQueuedClaudeStreamWriter = (write) => {
 		return pending;
 	};
 };
-const startClaudeStreamKeepAlive = (writeEvent) => {
-	const writePing = () => {
-		writeEvent({ type: "ping" }).catch(() => void 0);
-	};
-	writePing();
-	const timer = setInterval(writePing, streamKeepAliveIntervalMs);
-	if (typeof timer.unref === "function") timer.unref();
-	return () => {
-		clearInterval(timer);
-	};
-};
 const writeClaudeStreamEvents = async (events$1, writeEvent) => {
 	for (const event of events$1) await writeEvent(event);
 };
@@ -2667,14 +2655,12 @@ claudeRoutes.post("/messages", async (c) => {
 			event: event.type,
 			data: JSON.stringify(event)
 		}));
-		const stopKeepAlive = startClaudeStreamKeepAlive(writeEvent);
 		try {
 			await handleClaudeMessageRequest(config, claudePayload, requestSignal, requestId, writeEvent);
 		} catch (error) {
 			log.error("Error during Claude stream request:", error);
 			await writeEvent(translateErrorToClaudeErrorEvent());
 		} finally {
-			stopKeepAlive();
 			log.info(`request_id=${requestId} stream completed ${Math.round(performance.now() - streamStarted)}ms`);
 		}
 	});