@dianshuv/copilot-api 0.4.1 → 0.4.3

package/README.md

@@ -61,10 +61,11 @@ copilot-api start
 | `--proxy-env` | Use proxy from environment | false |
 | `--no-history` | Disable request history UI at `/history` | false |
 | `--history-limit` | Max history entries in memory | 1000 |
-| `--no-auto-truncate` | Disable auto-truncate when exceeding limits | false |
+| `--no-auto-truncate` | Disable auto-truncate when exceeding token limits | false |
 | `--compress-tool-results` | Compress old tool results before truncating | false |
 | `--redirect-anthropic` | Force Anthropic through OpenAI translation | false |
 | `--no-rewrite-anthropic-tools` | Don't rewrite server-side tools | false |
+| `--timezone-offset` | Timezone offset in hours from UTC for log timestamps (e.g., +8, -5, 0) | +8 |
 
 ### Patch-Claude Command Options
 
@@ -105,6 +106,15 @@ copilot-api start
 | `/history` | GET | Request history Web UI with token analytics (enabled by default) |
 | `/history/api/*` | GET/DELETE | History API endpoints |
 
+## Auto-Truncate
+
+When enabled (default), auto-truncate automatically compacts conversation history when it exceeds the model's token limit. This prevents request failures due to context overflow.
+
+- **Token-based truncation**: Uses the model's `max_context_window_tokens` from the Copilot API to determine when truncation is needed. A 2% safety margin is applied.
+- **No preset byte limit**: There is no hardcoded request body size limit. If the Copilot API returns a 413 (Request Entity Too Large), the proxy dynamically learns the byte limit and applies it to subsequent requests.
+- **Smart compression**: With `--compress-tool-results`, old tool results are compressed before removing messages, preserving more conversation context.
+- **Orphan filtering**: After truncation, orphaned tool results (without matching tool calls) are automatically removed.
+
 ## Using with Claude Code
 
 Create `.claude/settings.json` in your project:

package/dist/main.mjs

@@ -52,6 +52,7 @@ const state = {
 	redirectAnthropic: false,
 	rewriteAnthropicTools: true,
 	staleRequestMaxAge: 600,
+	timezoneOffset: 8,
 	shutdownGracefulWait: 60,
 	shutdownAbortWait: 120
 };
@@ -105,7 +106,7 @@ const GITHUB_APP_SCOPES = ["read:user"].join(" ");
 */
 const DEFAULT_AUTO_TRUNCATE_CONFIG = {
 	safetyMarginPercent: 2,
-	maxRequestBodyBytes: 510 * 1024,
+	maxRequestBodyBytes: Infinity,
 	preserveRecentPercent: .7
 };
 /** Dynamic byte limit that adjusts based on 413 errors */
@@ -1020,7 +1021,7 @@ const patchClaude = defineCommand({
 
 //#endregion
 //#region package.json
-var version = "0.4.1";
+var version = "0.4.3";
 
 //#endregion
 //#region src/lib/adaptive-rate-limiter.ts
@@ -2199,7 +2200,8 @@ function setupShutdownHandlers() {
 //#region src/lib/tui/console-renderer.ts
 const CLEAR_LINE = "\x1B[2K\r";
 function formatTime(date = /* @__PURE__ */ new Date()) {
-	return `${String(date.getHours()).padStart(2, "0")}:${String(date.getMinutes()).padStart(2, "0")}:${String(date.getSeconds()).padStart(2, "0")}`;
+	const adjusted = new Date(date.getTime() + state.timezoneOffset * 60 * 60 * 1e3);
+	return `${String(adjusted.getUTCHours()).padStart(2, "0")}:${String(adjusted.getUTCMinutes()).padStart(2, "0")}:${String(adjusted.getUTCSeconds()).padStart(2, "0")}`;
 }
 function formatDuration(ms) {
 	if (ms < 1e3) return `${ms}ms`;
@@ -3519,7 +3521,7 @@ async function buildFinalPayload(payload, model) {
 	}
 	try {
 		const check = await checkNeedsCompactionOpenAI(payload, model);
-		consola.debug(`Auto-truncate check: ${check.currentTokens} tokens (limit ${check.tokenLimit}), ${Math.round(check.currentBytes / 1024)}KB (limit ${Math.round(check.byteLimit / 1024)}KB), needed: ${check.needed}${check.reason ? ` (${check.reason})` : ""}`);
+		consola.debug(`Auto-truncate check: ${check.currentTokens} tokens (limit ${check.tokenLimit}), ${Math.round(check.currentBytes / 1024)}KB (limit ${check.byteLimit === Infinity ? "unlimited" : `${Math.round(check.byteLimit / 1024)}KB`}), needed: ${check.needed}${check.reason ? ` (${check.reason})` : ""}`);
 		if (!check.needed) return {
 			finalPayload: payload,
 			truncateResult: null
@@ -6736,7 +6738,7 @@ async function handleDirectAnthropicCompletion(c, anthropicPayload, ctx, initiat
 	let truncateResult;
 	if (state.autoTruncate && selectedModel) {
 		const check = await checkNeedsCompactionAnthropic(anthropicPayload, selectedModel);
-		consola.debug(`[Anthropic] Auto-truncate check: ${check.currentTokens} tokens (limit ${check.tokenLimit}), ${Math.round(check.currentBytes / 1024)}KB (limit ${Math.round(check.byteLimit / 1024)}KB), needed: ${check.needed}${check.reason ? ` (${check.reason})` : ""}`);
+		consola.debug(`[Anthropic] Auto-truncate check: ${check.currentTokens} tokens (limit ${check.tokenLimit}), ${Math.round(check.currentBytes / 1024)}KB (limit ${check.byteLimit === Infinity ? "unlimited" : `${Math.round(check.byteLimit / 1024)}KB`}), needed: ${check.needed}${check.reason ? ` (${check.reason})` : ""}`);
 		if (check.needed) try {
 			truncateResult = await autoTruncateAnthropic(anthropicPayload, selectedModel);
 			if (truncateResult.wasCompacted) effectivePayload = truncateResult.payload;
@@ -7815,6 +7817,7 @@ async function runServer(options) {
 	state.compressToolResults = options.compressToolResults;
 	state.redirectAnthropic = options.redirectAnthropic;
 	state.rewriteAnthropicTools = options.rewriteAnthropicTools;
+	state.timezoneOffset = options.timezoneOffset;
 	if (options.rateLimit) initAdaptiveRateLimiter({
 		baseRetryIntervalSeconds: options.retryInterval,
 		requestIntervalSeconds: options.requestInterval,
@@ -7879,6 +7882,12 @@ async function runServer(options) {
 		hostname: options.host
 	}));
 }
+function parseTimezoneOffset(value) {
+	if (typeof value !== "string") return 8;
+	const n = Number(value);
+	if (!Number.isFinite(n)) return 8;
+	return n;
+}
 const start = defineCommand({
 	meta: {
 		name: "start",
@@ -7988,6 +7997,11 @@ const start = defineCommand({
 			type: "boolean",
 			default: false,
 			description: "Don't rewrite Anthropic server-side tools (web_search, etc.) to custom tool format"
+		},
+		"timezone-offset": {
+			type: "string",
+			default: "+8",
+			description: "Timezone offset in hours from UTC for log timestamps (e.g., +8, -5, 0)"
 		}
 	},
 	run({ args }) {
@@ -8011,7 +8025,8 @@ const start = defineCommand({
 			autoTruncate: !args["no-auto-truncate"],
 			compressToolResults: args["compress-tool-results"],
 			redirectAnthropic: args["redirect-anthropic"],
-			rewriteAnthropicTools: !args["no-rewrite-anthropic-tools"]
+			rewriteAnthropicTools: !args["no-rewrite-anthropic-tools"],
+			timezoneOffset: parseTimezoneOffset(args["timezone-offset"])
 		});
 	}
 });

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@dianshuv/copilot-api",
-  "version": "0.4.1",
+  "version": "0.4.3",
   "description": "Turn GitHub Copilot into OpenAI/Anthropic API compatible server. Usable with Claude Code!",
   "author": "dianshuv",
   "type": "module",