@psnext/slingcli 2.5.20260623-1 → 2.5.20260629-1

package/node_modules/@earendil-works/pi-ai/dist/providers/xai.models.js

@@ -7,6 +7,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: false,
         input: ["text"],
         cost: {
@@ -24,6 +25,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: false,
         input: ["text"],
         cost: {
@@ -41,6 +43,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: false,
         input: ["text", "image"],
         cost: {
@@ -58,6 +61,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: true,
         input: ["text", "image"],
         cost: {
@@ -75,6 +79,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: true,
         input: ["text", "image"],
         cost: {
@@ -92,6 +97,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: true,
         input: ["text", "image"],
         cost: {
@@ -109,6 +115,7 @@ export const XAI_MODELS = {
         api: "openai-completions",
         provider: "xai",
         baseUrl: "https://api.x.ai/v1",
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false },
         reasoning: false,
         input: ["text"],
         cost: {

package/node_modules/@earendil-works/pi-ai/dist/providers/zai-coding-cn.models.js

@@ -7,7 +7,7 @@ export const ZAI_CODING_CN_MODELS = {
         api: "openai-completions",
         provider: "zai-coding-cn",
         baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai" },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai" },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -25,7 +25,7 @@ export const ZAI_CODING_CN_MODELS = {
         api: "openai-completions",
         provider: "zai-coding-cn",
         baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -43,7 +43,7 @@ export const ZAI_CODING_CN_MODELS = {
         api: "openai-completions",
         provider: "zai-coding-cn",
         baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -61,7 +61,7 @@ export const ZAI_CODING_CN_MODELS = {
         api: "openai-completions",
         provider: "zai-coding-cn",
         baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -79,7 +79,7 @@ export const ZAI_CODING_CN_MODELS = {
         api: "openai-completions",
         provider: "zai-coding-cn",
         baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "supportsReasoningEffort": true, "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": true, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         thinkingLevelMap: { "minimal": null, "low": "high", "medium": "high", "high": "high", "xhigh": "max" },
         input: ["text"],
@@ -98,7 +98,7 @@ export const ZAI_CODING_CN_MODELS = {
         api: "openai-completions",
         provider: "zai-coding-cn",
         baseUrl: "https://open.bigmodel.cn/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text", "image"],
         cost: {

package/node_modules/@earendil-works/pi-ai/dist/providers/zai.models.js

@@ -7,7 +7,7 @@ export const ZAI_MODELS = {
         api: "openai-completions",
         provider: "zai",
         baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai" },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai" },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -25,7 +25,7 @@ export const ZAI_MODELS = {
         api: "openai-completions",
         provider: "zai",
         baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -43,7 +43,7 @@ export const ZAI_MODELS = {
         api: "openai-completions",
         provider: "zai",
         baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -61,7 +61,7 @@ export const ZAI_MODELS = {
         api: "openai-completions",
         provider: "zai",
         baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text"],
         cost: {
@@ -79,7 +79,7 @@ export const ZAI_MODELS = {
         api: "openai-completions",
         provider: "zai",
         baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "supportsReasoningEffort": true, "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": true, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         thinkingLevelMap: { "minimal": null, "low": "high", "medium": "high", "high": "high", "xhigh": "max" },
         input: ["text"],
@@ -98,7 +98,7 @@ export const ZAI_MODELS = {
         api: "openai-completions",
         provider: "zai",
         baseUrl: "https://api.z.ai/api/coding/paas/v4",
-        compat: { "supportsDeveloperRole": false, "thinkingFormat": "zai", "zaiToolStream": true },
+        compat: { "supportsStore": false, "supportsDeveloperRole": false, "supportsReasoningEffort": false, "thinkingFormat": "zai", "zaiToolStream": true },
         reasoning: true,
         input: ["text", "image"],
         cost: {

package/node_modules/@earendil-works/pi-ai/dist/utils/estimate.js

@@ -0,0 +1,95 @@
+const CHARS_PER_TOKEN = 4;
+const ESTIMATED_IMAGE_CHARS = 4800;
+export function calculateContextTokens(usage) {
+    return usage.totalTokens || usage.input + usage.output + usage.cacheRead + usage.cacheWrite;
+}
+function safeJsonStringify(value) {
+    try {
+        return JSON.stringify(value) ?? "undefined";
+    }
+    catch {
+        return "[unserializable]";
+    }
+}
+function estimateTextAndImageContentChars(content) {
+    if (typeof content === "string")
+        return content.length;
+    let chars = 0;
+    for (const block of content)
+        chars += block.type === "text" ? block.text.length : ESTIMATED_IMAGE_CHARS;
+    return chars;
+}
+export function estimateTextTokens(text) {
+    return Math.ceil(text.length / CHARS_PER_TOKEN);
+}
+export function estimateTextAndImageContentTokens(content) {
+    return Math.ceil(estimateTextAndImageContentChars(content) / CHARS_PER_TOKEN);
+}
+export function estimateMessageTokens(message) {
+    let chars = 0;
+    if (message.role === "user")
+        return estimateTextAndImageContentTokens(message.content);
+    if (message.role === "toolResult")
+        return estimateTextAndImageContentTokens(message.content);
+    for (const block of message.content) {
+        if (block.type === "text") {
+            chars += block.text.length;
+        }
+        else if (block.type === "thinking") {
+            chars += block.thinking.length;
+        }
+        else {
+            chars += block.name.length + safeJsonStringify(block.arguments).length;
+        }
+    }
+    return Math.ceil(chars / CHARS_PER_TOKEN);
+}
+function getLastAssistantUsageInfo(messages) {
+    for (let i = messages.length - 1; i >= 0; i--) {
+        const message = messages[i];
+        if (message.role !== "assistant")
+            continue;
+        const assistant = message;
+        if (assistant.stopReason === "aborted" || assistant.stopReason === "error")
+            continue;
+        if (calculateContextTokens(assistant.usage) > 0)
+            return { usage: assistant.usage, index: i };
+    }
+    return undefined;
+}
+function estimateMessages(messages) {
+    const usageInfo = getLastAssistantUsageInfo(messages);
+    if (usageInfo) {
+        const usageTokens = calculateContextTokens(usageInfo.usage);
+        let trailingTokens = 0;
+        for (let i = usageInfo.index + 1; i < messages.length; i++) {
+            trailingTokens += estimateMessageTokens(messages[i]);
+        }
+        return { tokens: usageTokens + trailingTokens, usageTokens, trailingTokens, lastUsageIndex: usageInfo.index };
+    }
+    let tokens = 0;
+    for (const message of messages)
+        tokens += estimateMessageTokens(message);
+    return { tokens, usageTokens: 0, trailingTokens: tokens, lastUsageIndex: null };
+}
+function isMessageArray(value) {
+    return Array.isArray(value);
+}
+export function estimateContextTokens(context) {
+    if (isMessageArray(context))
+        return estimateMessages(context);
+    const estimate = estimateMessages(context.messages);
+    if (estimate.lastUsageIndex !== null)
+        return estimate;
+    let prefixTokens = context.systemPrompt ? estimateTextTokens(context.systemPrompt) : 0;
+    if (context.tools && context.tools.length > 0) {
+        prefixTokens += estimateTextTokens(safeJsonStringify(context.tools));
+    }
+    return {
+        tokens: estimate.tokens + prefixTokens,
+        usageTokens: estimate.usageTokens,
+        trailingTokens: estimate.trailingTokens + prefixTokens,
+        lastUsageIndex: estimate.lastUsageIndex,
+    };
+}
+//# sourceMappingURL=estimate.js.map

package/node_modules/@earendil-works/pi-ai/dist/utils/headers.js

@@ -5,4 +5,14 @@ export function headersToRecord(headers) {
     }
     return result;
 }
+export function providerHeadersToRecord(headers) {
+    if (!headers)
+        return undefined;
+    const result = {};
+    for (const [key, value] of Object.entries(headers)) {
+        if (value !== null)
+            result[key] = value;
+    }
+    return Object.keys(result).length > 0 ? result : undefined;
+}
 //# sourceMappingURL=headers.js.map

package/node_modules/@earendil-works/pi-ai/dist/utils/oauth/openai-codex.js

@@ -323,9 +323,9 @@ function getAccountId(accessToken) {
     return typeof accountId === "string" && accountId.length > 0 ? accountId : null;
 }
 function credentialsFromToken(token) {
-    const accountId = getAccountId(token.access) || "unknown";
+    const accountId = getAccountId(token.access);
     if (!accountId) {
-        console.log("Failed to extract accountId from token");
+        throw new Error("Failed to extract accountId from token");
     }
     return {
         access: token.access,

package/node_modules/@earendil-works/pi-ai/dist/utils/retry.js

@@ -0,0 +1,86 @@
+function buildProviderErrorPattern(patterns) {
+    return new RegExp(patterns.join("|"), "i");
+}
+const NON_RETRYABLE_PROVIDER_LIMIT_ERROR_PATTERN = buildProviderErrorPattern([
+    // OpenCode Go/free-tier limits returned as 429 JSON error types by OpenCode's
+    // Zen API. These are subscription/account limits, not transient throttles.
+    "GoUsageLimitError",
+    "FreeUsageLimitError",
+    // OpenCode Go subscription-limit text asks users to enable available-balance
+    // usage after rolling/weekly/monthly limits are reached.
+    "Monthly usage limit reached",
+    "available balance",
+    // Generic quota/budget/billing exhaustion. `insufficient_quota` is OpenAI's
+    // quota/billing error code; the other strings cover common gateway wording.
+    "insufficient_quota",
+    "out of budget",
+    "quota exceeded",
+    "billing",
+]);
+const RETRYABLE_PROVIDER_ERROR_PATTERN = buildProviderErrorPattern([
+    // Generic provider load, HTTP status, and server-side transient failures.
+    "overloaded",
+    "rate.?limit",
+    "too many requests",
+    "429",
+    "500",
+    "502",
+    "503",
+    "504",
+    "service.?unavailable",
+    "server.?error",
+    "internal.?error",
+    // Wrapper/provider text for transient upstream failures, including OpenRouter
+    // "Provider returned error" responses (#2264).
+    "provider.?returned.?error",
+    // Network, proxy, and fetch transport failures. This includes OpenAI Codex
+    // raw-fetch failures such as "upstream connect", "connection refused", and
+    // "reset before headers" (#733), plus OpenRouter connection drops (#3317).
+    "network.?error",
+    "connection.?error",
+    "connection.?refused",
+    "connection.?lost",
+    "other side closed",
+    "fetch failed",
+    "upstream.?connect",
+    "reset before headers",
+    "socket hang up",
+    "timed? out",
+    "timeout",
+    "terminated",
+    // WebSocket transports can report close/error text instead of HTTP/fetch text.
+    "websocket.?closed",
+    "websocket.?error",
+    // Premature stream endings from SDKs and transports. Anthropic can throw
+    // "stream ended without ..." and "Anthropic stream ended before message_stop"
+    // (#4433); Bedrock/Smithy can throw an HTTP/2 no-response error (#3594).
+    "ended without",
+    "stream ended before message_stop",
+    "http2 request did not get a response",
+    // Provider-requested retry delay cap failures should flow through the outer
+    // retry policy so callers can surface/abort the backoff (#1123).
+    "retry delay",
+    // Explicit retry guidance emitted mid-stream by OpenAI Responses and Bedrock
+    // stream exceptions (#6019).
+    "you can retry your request",
+    "try your request again",
+    "please retry your request",
+]);
+/**
+ * Classifies whether a failed assistant message looks like a transient provider
+ * or transport error, so callers can decide if the last assistant turn should be
+ * restarted.
+ *
+ * This does not implement retry policy. Callers should first handle context
+ * overflow separately, then apply their own retry budget, backoff, and reporting
+ * before restarting the assistant turn.
+ */
+export function isRetryableAssistantError(message) {
+    if (message.stopReason !== "error" || !message.errorMessage)
+        return false;
+    const errorMessage = message.errorMessage;
+    if (NON_RETRYABLE_PROVIDER_LIMIT_ERROR_PATTERN.test(errorMessage))
+        return false;
+    return RETRYABLE_PROVIDER_ERROR_PATTERN.test(errorMessage);
+}
+//# sourceMappingURL=retry.js.map

package/node_modules/@earendil-works/pi-ai/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@earendil-works/pi-ai",
-	"version": "0.79.10",
+	"version": "0.80.2",
 	"description": "Unified LLM API with automatic model discovery and provider configuration",
 	"type": "module",
 	"main": "./dist/index.js",
@@ -27,38 +27,6 @@
 			"types": "./dist/api/*.d.ts",
 			"import": "./dist/api/*.js"
 		},
-		"./anthropic": {
-			"types": "./dist/api/anthropic-messages.d.ts",
-			"import": "./dist/api/anthropic-messages.js"
-		},
-		"./azure-openai-responses": {
-			"types": "./dist/api/azure-openai-responses.d.ts",
-			"import": "./dist/api/azure-openai-responses.js"
-		},
-		"./google": {
-			"types": "./dist/api/google-generative-ai.d.ts",
-			"import": "./dist/api/google-generative-ai.js"
-		},
-		"./google-vertex": {
-			"types": "./dist/api/google-vertex.d.ts",
-			"import": "./dist/api/google-vertex.js"
-		},
-		"./mistral": {
-			"types": "./dist/api/mistral-conversations.d.ts",
-			"import": "./dist/api/mistral-conversations.js"
-		},
-		"./openai-codex-responses": {
-			"types": "./dist/api/openai-codex-responses.d.ts",
-			"import": "./dist/api/openai-codex-responses.js"
-		},
-		"./openai-completions": {
-			"types": "./dist/api/openai-completions.d.ts",
-			"import": "./dist/api/openai-completions.js"
-		},
-		"./openai-responses": {
-			"types": "./dist/api/openai-responses.d.ts",
-			"import": "./dist/api/openai-responses.js"
-		},
 		"./oauth": {
 			"types": "./dist/oauth.d.ts",
 			"import": "./dist/oauth.js"

package/node_modules/@earendil-works/pi-coding-agent/dist/cli/file-processor.js

@@ -5,7 +5,7 @@ import { access, readFile, stat } from "node:fs/promises";
 import chalk from "chalk";
 import { resolve } from "path";
 import { resolveReadPath } from "../core/tools/path-utils.js";
-import { formatDimensionNote, resizeImage } from "../utils/image-resize.js";
+import { processImage } from "../utils/image-process.js";
 import { detectSupportedImageMimeTypeFromFile } from "../utils/mime.js";
 /** Process @file arguments into text content and image attachments */
 export async function processFileArguments(fileArgs, options) {
@@ -33,32 +33,20 @@ export async function processFileArguments(fileArgs, options) {
         if (mimeType) {
             // Handle image file
             const content = await readFile(absolutePath);
-            let attachment;
-            let dimensionNote;
-            if (autoResizeImages) {
-                const resized = await resizeImage(content, mimeType);
-                if (!resized) {
-                    text += `<file name="${absolutePath}">[Image omitted: could not be resized below the inline image size limit.]</file>\n`;
-                    continue;
-                }
-                dimensionNote = formatDimensionNote(resized);
-                attachment = {
-                    type: "image",
-                    mimeType: resized.mimeType,
-                    data: resized.data,
-                };
-            }
-            else {
-                attachment = {
-                    type: "image",
-                    mimeType,
-                    data: content.toString("base64"),
-                };
+            const processed = await processImage(content, mimeType, { autoResizeImages });
+            if (!processed.ok) {
+                text += `<file name="${absolutePath}">${processed.message}</file>\n`;
+                continue;
             }
+            const attachment = {
+                type: "image",
+                mimeType: processed.mimeType,
+                data: processed.data,
+            };
             images.push(attachment);
-            // Add text reference to image with optional dimension note
-            if (dimensionNote) {
-                text += `<file name="${absolutePath}">${dimensionNote}</file>\n`;
+            // Add text reference to image with optional processing hints
+            if (processed.hints.length > 0) {
+                text += `<file name="${absolutePath}">${processed.hints.join("\n")}</file>\n`;
             }
             else {
                 text += `<file name="${absolutePath}"></file>\n`;

package/node_modules/@earendil-works/pi-coding-agent/dist/core/agent-session.js

@@ -14,7 +14,7 @@
  */
 import { existsSync, mkdirSync, readFileSync, writeFileSync } from "node:fs";
 import { basename, dirname } from "node:path";
-import { clampThinkingLevel, cleanupSessionResources, getSupportedThinkingLevels, isContextOverflow, modelsAreEqual, resetApiProviders, streamSimple, } from "@earendil-works/pi-ai/compat";
+import { clampThinkingLevel, cleanupSessionResources, getSupportedThinkingLevels, isContextOverflow, isRetryableAssistantError, modelsAreEqual, resetApiProviders, streamSimple, } from "@earendil-works/pi-ai/compat";
 import { getThemeByName, theme } from "../modes/interactive/theme/theme.js";
 import { stripFrontmatter } from "../utils/frontmatter.js";
 import { resolvePath } from "../utils/paths.js";
@@ -769,18 +769,11 @@ export class AgentSession {
                 }
                 throw new Error(formatNoApiKeyFoundMessage(this.model.provider));
             }
-            // Check if we need to compact before sending (catches aborted responses)
+            // Check if we need to compact before sending (catches aborted responses).
+            // The user's new prompt is sent below, so do not call agent.continue() here.
             const lastAssistant = this._findLastAssistantMessage();
-            if (lastAssistant && (await this._checkCompaction(lastAssistant, false))) {
-                try {
-                    await this.agent.continue();
-                    while (await this._handlePostAgentRun()) {
-                        await this.agent.continue();
-                    }
-                }
-                finally {
-                    this._flushPendingBashMessages();
-                }
+            if (lastAssistant) {
+                await this._checkCompaction(lastAssistant, false);
             }
             // Build messages array (custom message if any, then user message)
             messages = [];
@@ -1459,10 +1452,12 @@ export class AgentSession {
             return await this._runAutoCompaction("overflow", willRetry);
         }
         // Case 2: Threshold - context is getting large
-        // For error messages (no usage data), estimate from last successful response.
-        // This ensures sessions that hit persistent API errors (e.g. 529) can still compact.
+        // For error messages or all-zero usage messages, estimate from the last valid response.
+        // This ensures sessions that hit persistent API errors (e.g. 529) or malformed zero-usage
+        // responses can still compact and do not reset context accounting.
         let contextTokens;
-        if (assistantMessage.stopReason === "error") {
+        const directContextTokens = assistantMessage.usage ? calculateContextTokens(assistantMessage.usage) : 0;
+        if (assistantMessage.stopReason === "error" || directContextTokens === 0) {
             const messages = this.agent.state.messages;
             const estimate = estimateContextTokens(messages);
             if (estimate.lastUsageIndex === null)
@@ -1479,7 +1474,7 @@ export class AgentSession {
             contextTokens = estimate.tokens;
         }
         else {
-            contextTokens = calculateContextTokens(assistantMessage.usage);
+            contextTokens = directContextTokens;
         }
         if (shouldCompact(contextTokens, contextWindow, settings)) {
             return await this._runAutoCompaction("threshold", false);
@@ -1968,25 +1963,15 @@ export class AgentSession {
     // =========================================================================
     // Auto-Retry
     // =========================================================================
-    _isNonRetryableProviderLimitError(errorMessage) {
-        return /GoUsageLimitError|FreeUsageLimitError|Monthly usage limit reached|available balance|insufficient_quota|out of budget|quota exceeded|billing/i.test(errorMessage);
-    }
     /**
      * Check if an error is retryable (overloaded, rate limit, server errors).
      * Context overflow errors are NOT retryable (handled by compaction instead).
      */
     _isRetryableError(message) {
-        if (message.stopReason !== "error" || !message.errorMessage)
-            return false;
-        // Context overflow is handled by compaction, not retry
-        const contextWindow = this.model?.contextWindow ?? 0;
-        if (isContextOverflow(message, contextWindow))
-            return false;
-        const err = message.errorMessage;
-        if (this._isNonRetryableProviderLimitError(err))
+        // Context overflow is handled by compaction, not retry.
+        if (isContextOverflow(message, this.model?.contextWindow ?? 0))
             return false;
-        // Match: overloaded_error, provider returned error, rate limit, 429, 500, 502, 503, 504, service unavailable, network/connection errors (including connection lost), WebSocket transport closes/errors, fetch failed, premature stream endings, HTTP/2 closed before response, terminated, retry delay exceeded
-        return /overloaded|provider.?returned.?error|rate.?limit|too many requests|429|500|502|503|504|service.?unavailable|server.?error|internal.?error|network.?error|connection.?error|connection.?refused|connection.?lost|websocket.?closed|websocket.?error|other side closed|fetch failed|upstream.?connect|reset before headers|socket hang up|ended without|stream ended before message_stop|http2 request did not get a response|timed? out|timeout|terminated|retry delay/i.test(err);
+        return isRetryableAssistantError(message);
     }
     /**
      * Prepare a retryable error for continuation with exponential backoff.
@@ -2424,8 +2409,8 @@ export class AgentSession {
                         const contextTokens = calculateContextTokens(assistant.usage);
                         if (contextTokens > 0) {
                             hasPostCompactionUsage = true;
+                            break;
                         }
-                        break;
                     }
                 }
             }

package/node_modules/@earendil-works/pi-coding-agent/dist/core/compaction/compaction.js

@@ -80,19 +80,22 @@ export function calculateContextTokens(usage) {
 }
 /**
  * Get usage from an assistant message if available.
- * Skips aborted and error messages as they don't have valid usage data.
+ * Skips aborted, error, and all-zero usage messages as they don't have valid usage data.
  */
 function getAssistantUsage(msg) {
     if (msg.role === "assistant" && "usage" in msg) {
         const assistantMsg = msg;
-        if (assistantMsg.stopReason !== "aborted" && assistantMsg.stopReason !== "error" && assistantMsg.usage) {
+        if (assistantMsg.stopReason !== "aborted" &&
+            assistantMsg.stopReason !== "error" &&
+            assistantMsg.usage &&
+            calculateContextTokens(assistantMsg.usage) > 0) {
             return assistantMsg.usage;
         }
     }
     return undefined;
 }
 /**
- * Find the last non-aborted assistant message usage from session entries.
+ * Find the last valid assistant message usage from session entries.
  */
 export function getLastAssistantUsage(entries) {
     for (let i = entries.length - 1; i >= 0; i--) {

package/node_modules/@earendil-works/pi-coding-agent/dist/core/extensions/loader.js

@@ -25,6 +25,7 @@ import { resolvePath } from "../../utils/paths.js";
 import { createEventBus } from "../event-bus.js";
 import { execCommand } from "../exec.js";
 import { createSyntheticSourceInfo } from "../source-info.js";
+import { time } from "../timings.js";
 /** Modules available to extensions via virtualModules (for compiled Bun binary) */
 const VIRTUAL_MODULES = {
     typebox: _bundledTypebox,
@@ -153,7 +154,7 @@ export function createExtensionRuntime() {
         invalidate: (message) => {
             state.staleMessage ??=
                 message ??
-                    "This extension ctx is stale after session replacement or reload. Do not use a captured sling or command ctx after ctx.newSession(), ctx.fork(), ctx.switchSession(), or ctx.reload(). For newSession, fork, and switchSession, move post-replacement work into withSession and use the ctx passed to withSession. For reload, do not use the old ctx after await ctx.reload().";
+                    "This extension ctx is stale after session replacement or reload. Do not use a captured pi or command ctx after ctx.newSession(), ctx.fork(), ctx.switchSession(), or ctx.reload(). For newSession, fork, and switchSession, move post-replacement work into withSession and use the ctx passed to withSession. For reload, do not use the old ctx after await ctx.reload().";
         },
         // Pre-bind: queue registrations so bindCore() can flush them once the
         // model registry is available. bindCore() replaces both with direct calls.
@@ -340,12 +341,14 @@ async function loadExtension(extensionPath, cwd, eventBus, runtime, cacheToken)
     const resolvedPath = resolvePath(extensionPath, cwd, { normalizeUnicodeSpaces: true });
     try {
         const factory = await loadExtensionModule(resolvedPath, cacheToken);
+        time(`${extensionPath} module import`, "extensions");
         if (!factory) {
             return { extension: null, error: `Extension does not export a valid factory function: ${extensionPath}` };
         }
         const extension = createExtension(extensionPath, resolvedPath);
         const api = createExtensionAPI(extension, runtime, cwd, eventBus);
         await factory(api);
+        time(`${extensionPath} factory`, "extensions");
         return { extension, error: null };
     }
     catch (err) {
@@ -361,6 +364,7 @@ export async function loadExtensionFromFactory(factory, cwd, eventBus, runtime,
     const resolvedCwd = resolvePath(cwd);
     const api = createExtensionAPI(extension, runtime, resolvedCwd, eventBus);
     await factory(api);
+    time(`${extensionPath} factory`, "extensions");
     return extension;
 }
 /**

package/node_modules/@earendil-works/pi-coding-agent/dist/core/model-registry.js

@@ -402,13 +402,11 @@ export class ModelRegistry {
                 }
             }
             else if (!isBuiltIn) {
-                // Non-built-in providers with custom models require endpoint + auth.
+                // Non-built-in providers with custom models require an endpoint.
+                // Auth can come from auth.json, --api-key, or provider request config.
                 if (!providerConfig.baseUrl) {
                     throw new Error(`Provider ${providerName}: "baseUrl" is required when defining custom models.`);
                 }
-                if (!providerConfig.apiKey) {
-                    throw new Error(`Provider ${providerName}: "apiKey" is required when defining custom models.`);
-                }
             }
             // Built-in providers with custom models: baseUrl/apiKey/api are optional,
             // inherited from built-in models. Auth comes from env vars / auth storage.