kimiflare 0.55.0 → 0.56.0

package/README.md

@@ -33,7 +33,12 @@
 
 - **262k context window** — Read entire modules, large configs, and full stack traces without the model losing track.
 - **Image understanding** — Drop image paths (PNG, JPG, WebP, GIF, BMP up to 5 MB) into any prompt. Great for UI reviews, diagrams, and screenshots.
-- **Plan / Edit / Auto modes** — `plan` blocks mutating tools for safe research. `edit` (default) prompts per mutating call. `auto` approves everything for trusted tasks.
+- **Plan / Edit / Auto modes** — `plan` is a whitelist-only research mode: only read-only tools (read, glob, grep, web search, GitHub read-only, browser fetch) are allowed. Writes, edits, mutating bash, MCP tools, and LSP renames are all blocked. `edit` (default) prompts per mutating call. `auto` approves everything for trusted tasks.
+- **Windows support** — OS-aware shell auto-detects `cmd.exe` / PowerShell on Windows, `bash` on Unix. The `bash` tool works out of the box on all platforms.
+- **Message queuing** — Submit multiple messages while the agent is busy; they queue and auto-drain. Escape interrupts the current turn but preserves the queue.
+- **Smart permission modal** — Denying a tool opens inline feedback so you can tell the agent what to do instead. Keyboard-native navigation (`↑/↓`, `j/k`, `Alt+1/2/3`).
+- **Loop guardrails** — Agent hard-stops when all tools in a turn are blocked, preventing infinite token-burning cycles.
+- **Persistent all-time cost history** — Append-only `history.jsonl` tracks daily usage forever, so `/cost` shows true all-time and monthly totals that survive across sessions and version updates.
 - **Live cost tracking** — Status bar shows real-time spend based on Cloudflare pricing. Know exactly what each turn costs.
 - **LSP + MCP** — Semantic code intelligence (hover, go-to-definition, references, diagnostics) via Language Server Protocol. Extend with external tools via Model Context Protocol.
 - **Local structured memory** — SQLite + embeddings cross-session memory. The agent recalls facts, instructions, and preferences across sessions via `remember`, `recall`, and `forget` tools.
@@ -41,14 +46,16 @@
 
 ## Recently shipped
 
-- **Turn supervisor architecture** — graceful preemption, visual cleanup, and better multi-step task management.
-- **Web search, GitHub read-only, and headless browser tools** — research without leaving the terminal.
-- **Tiered skill routing** — the agent picks the right skill depth for the task, with visible TUI indicators.
-- **Extensible JSON themes** — WCAG contrast-validated, fully customizable color palettes.
-- **KIMI.md drift detection** — memory-based staleness indicators warn when your project context file is out of date.
-- **Fuzzy @ file picker** — type `@` to mention files with fuzzy matching and inline filtering.
-- **Kimiflare Cloud mode** — device auth, no API key needed, with real-time token budget tracking.
-- **Context-window guardrails** — prevents runs that would exceed the model's limit before they start.
+- **OS-aware shell with Windows support** — Auto-detects `cmd.exe`, PowerShell, or bash based on platform. Override with `KIMIFLARE_SHELL` or `/shell`.
+- **Smart permission modal with inline feedback** — Deny a tool and immediately tell the agent what to do instead. Keyboard-native navigation with `↑/↓`, `j/k`, `Alt+1/2/3`.
+- **True message queuing** — Enter queues messages while the agent is busy; Escape interrupts and auto-drains the queue.
+- **Hard-stop loop guardrail** — Stops token-burning cycles when all tools in a turn are blocked.
+- **Persistent all-time usage history** — `history.jsonl` tracks daily usage forever; `/cost` shows true all-time and monthly totals.
+- **Humanized Cloudflare API errors** — Actionable error codes and structured error display instead of raw JSON dumps.
+- **429 rate limit retry** — Automatic backoff and retry when Cloudflare rate-limits requests.
+- **Tool state visualization** — Queued, rejected, and cancelled tools are clearly labeled in the TUI.
+- **Paste preview placeholders** — Pasted content shows a snippet preview with sequential IDs instead of random hashes.
+- **Headless SDK** — Programmatic `createAgentSession` API and JSONL-over-stdio RPC mode for building on top of KimiFlare.
 
 See the full changelog at [github.com/sinameraji/kimiflare/releases](https://github.com/sinameraji/kimiflare/releases).
 
@@ -112,7 +119,7 @@ session.dispose();
 ```
 
 **Key features:**
-- `subscribe()` — receive typed events (`text_delta`, `tool_call`, `tool_result`, `task_update`, `usage`, `error`, `done`, etc.)
+- `subscribe()` — receive typed events (`text_delta`, `tool_call`, `tool_result`, `task_update`, `usage`, `warning`, `error`, `done`, etc.)
 - `prompt()` / `steer()` / `followUp()` — full conversation lifecycle
 - `pause()` / `resume()` — graceful preemption
 - `getStatus()` / `getUsage()` — inspect session state
@@ -191,6 +198,7 @@ kimiflare
 | Command | Effect |
 |---------|--------|
 | `/mode edit\|plan\|auto` | Switch permission mode |
+| `/shell auto\|bash\|cmd\|powershell` | Show or set the shell for the bash tool |
 | `/thinking low\|medium\|high` | Reasoning effort (persists) |
 | `/theme` | Interactive theme picker (`Ctrl+T`) |
 | `/resume` | Pick a past conversation to restore |

package/dist/index.js

@@ -140,7 +140,7 @@ async function loadConfig() {
       cacheStablePrompts,
       compiledContext,
       imageHistoryTurns: Number.isNaN(imageHistoryTurns) ? void 0 : imageHistoryTurns,
-      memoryEnabled: envMemoryEnabled,
+      memoryEnabled: envMemoryEnabled ?? true,
       memoryDbPath: envMemoryDbPath,
       memoryMaxAgeDays: envMemoryMaxAgeDays,
       memoryMaxEntries: envMemoryMaxEntries,
@@ -199,7 +199,7 @@ async function loadConfig() {
         cacheStablePrompts: parsed.cacheStablePrompts ?? cacheStablePrompts,
         compiledContext: parsed.compiledContext ?? compiledContext,
         imageHistoryTurns: Number.isNaN(imageHistoryTurns) ? parsed.imageHistoryTurns : imageHistoryTurns,
-        memoryEnabled: envMemoryEnabled ?? parsed.memoryEnabled,
+        memoryEnabled: envMemoryEnabled ?? parsed.memoryEnabled ?? true,
         memoryDbPath: envMemoryDbPath ?? parsed.memoryDbPath,
         memoryMaxAgeDays: envMemoryMaxAgeDays ?? parsed.memoryMaxAgeDays,
         memoryMaxEntries: envMemoryMaxEntries ?? parsed.memoryMaxEntries,
@@ -353,16 +353,21 @@ var init_lsp_config = __esm({
 
 // src/util/logger.ts
 function log(level, event, data) {
-  if (LEVEL_ORDER[level] < LEVEL_ORDER[globalMinLevel]) return;
   const entry = {
     ts: (/* @__PURE__ */ new Date()).toISOString(),
     level,
     event,
     data
   };
-  console.error(JSON.stringify(entry));
+  recentLogs.push(entry);
+  if (recentLogs.length > RECENT_LOGS_MAX) {
+    recentLogs.shift();
+  }
+  if (LEVEL_ORDER[level] >= LEVEL_ORDER[globalMinLevel]) {
+    console.error(JSON.stringify(entry));
+  }
 }
-var globalMinLevel, LEVEL_ORDER, logger;
+var globalMinLevel, LEVEL_ORDER, RECENT_LOGS_MAX, recentLogs, logger;
 var init_logger = __esm({
   "src/util/logger.ts"() {
     "use strict";
@@ -374,6 +379,8 @@ var init_logger = __esm({
       error: 3,
       off: 4
     };
+    RECENT_LOGS_MAX = 100;
+    recentLogs = [];
     logger = {
       debug: (event, data) => log("debug", event, data),
       info: (event, data) => log("info", event, data),
@@ -8866,7 +8873,7 @@ async function createAgentSession(opts2) {
   const tools = opts2.tools ?? ALL_TOOLS;
   const executor = new ToolExecutor(tools);
   let memoryManager = null;
-  const memoryEnabled = opts2.memoryEnabled ?? config.memoryEnabled ?? false;
+  const memoryEnabled = opts2.memoryEnabled ?? config.memoryEnabled ?? true;
   if (memoryEnabled) {
     const dbPath = config.memoryDbPath ?? join19(homedir12(), ".local", "share", "kimiflare", "memory.db");
     memoryManager = new MemoryManager({
@@ -9821,6 +9828,59 @@ var init_abort_scope = __esm({
   }
 });
 
+// src/cloud/report.ts
+function buildReport2(opts2) {
+  return {
+    error: {
+      message: opts2.errorMessage,
+      code: opts2.errorCode,
+      http_status: opts2.httpStatus
+    },
+    context: {
+      command: "/report",
+      model: opts2.model,
+      session_id: opts2.sessionId,
+      request_id: opts2.requestId
+    },
+    user_message: opts2.userNote,
+    metadata: {
+      version: getAppVersion(),
+      platform: `${process.platform} ${process.arch}`,
+      node_version: process.version,
+      cloud_mode: opts2.cloudMode ?? false
+    }
+  };
+}
+async function sendReport(payload, token) {
+  try {
+    const headers = { "Content-Type": "application/json" };
+    if (token) {
+      headers["Authorization"] = `Bearer ${token}`;
+    }
+    const res = await fetch(REPORT_URL, {
+      method: "POST",
+      headers,
+      body: JSON.stringify(payload)
+    });
+    if (res.ok) {
+      return { ok: true, message: "Report sent. Thanks for helping improve KimiFlare!" };
+    }
+    const body = await res.text().catch(() => "unknown error");
+    return { ok: false, message: `Failed to send report (${res.status}): ${body}` };
+  } catch (e) {
+    const msg = e instanceof Error ? e.message : String(e);
+    return { ok: false, message: `Failed to send report: ${msg}` };
+  }
+}
+var REPORT_URL;
+var init_report2 = __esm({
+  "src/cloud/report.ts"() {
+    "use strict";
+    init_version();
+    REPORT_URL = "https://api.kimiflare.com/v1/report";
+  }
+});
+
 // src/ui/theme-context.tsx
 import { createContext, useContext } from "react";
 import { jsx } from "react/jsx-runtime";
@@ -10629,7 +10689,8 @@ function ApiErrorMessage({ httpStatus, code, message: message2 }) {
       "\u26A0 ",
       message2
     ] }),
-    meta && /* @__PURE__ */ jsx6(Text5, { color: theme.muted?.color ?? theme.info.color, dimColor: theme.muted?.dim ?? true, children: meta })
+    meta && /* @__PURE__ */ jsx6(Text5, { color: theme.muted?.color ?? theme.info.color, dimColor: theme.muted?.dim ?? true, children: meta }),
+    /* @__PURE__ */ jsx6(Text5, { color: theme.muted?.color ?? theme.info.color, dimColor: theme.muted?.dim ?? true, children: "Type /report to send diagnostic info" })
   ] });
 }
 var init_api_error_message = __esm({
@@ -13766,6 +13827,7 @@ var init_builtins = __esm({
       { name: "remote", argHint: "<prompt>", description: "Run a remote session on Cloudflare", source: "builtin" },
       { name: "update", description: "Check for updates", source: "builtin" },
       { name: "hello", description: "Send a voice note to the creator", source: "builtin" },
+      { name: "report", argHint: "[send] [note]", description: "Report the last API error with diagnostic logs", source: "builtin" },
       { name: "shell", argHint: "[auto|bash|cmd|powershell|<path>]", description: "Show or set shell for bash tool", source: "builtin" },
       { name: "logout", description: "Clear stored credentials", source: "builtin" },
       { name: "exit", description: "Exit kimiflare", source: "builtin" },
@@ -17147,6 +17209,7 @@ ${wcagWarnings.join("\n")}` }
   const tasksRef = useRef3([]);
   const usageRef = useRef3(null);
   const gatewayMetaRef = useRef3(null);
+  const lastApiErrorRef = useRef3(null);
   const updateCheckedRef = useRef3(false);
   const sessionStateRef = useRef3(emptySessionState());
   const artifactStoreRef = useRef3(new ArtifactStore());
@@ -18368,15 +18431,11 @@ ${wcagWarnings.join("\n")}` }
           { kind: "error", key: mkKey(), text: "The agent got stuck repeating the same actions. Here's what we know so far." }
         ]);
       } else if (e instanceof KimiApiError && (e.httpStatus === 429 || e.code === 3040 || e.httpStatus !== void 0 && e.httpStatus >= 500)) {
+        const err = { httpStatus: e.httpStatus, code: e.code, message: humanizeCloudflareError(e) };
+        lastApiErrorRef.current = err;
         setEvents((es) => [
           ...es,
-          {
-            kind: "api_error",
-            key: mkKey(),
-            httpStatus: e.httpStatus,
-            code: e.code,
-            message: humanizeCloudflareError(e)
-          }
+          { kind: "api_error", key: mkKey(), ...err }
         ]);
       } else {
         const displayText = e instanceof KimiApiError ? humanizeCloudflareError(e) : `init failed: ${e.message}`;
@@ -19191,6 +19250,55 @@ ${lines.join("\n")}` }]);
         ]);
         return true;
       }
+      if (c === "/report") {
+        const err = lastApiErrorRef.current;
+        if (!err) {
+          setEvents((e) => [
+            ...e,
+            { kind: "info", key: mkKey(), text: "No recent API error to report." }
+          ]);
+          return true;
+        }
+        const note = rest.join(" ").trim();
+        const isSend = note.toLowerCase() === "send" || note.toLowerCase().startsWith("send ");
+        if (!isSend) {
+          const preview = [
+            "Report preview:",
+            `  Error: ${err.message}`,
+            err.httpStatus !== void 0 ? `  HTTP ${err.httpStatus}` : "",
+            err.code !== void 0 ? `  Code: ${err.code}` : "",
+            note ? `  Note: ${note}` : "",
+            "",
+            "Type `/report send` to submit or `/report send <note>` to add context."
+          ].filter(Boolean).join("\n");
+          setEvents((e) => [...e, { kind: "info", key: mkKey(), text: preview }]);
+          return true;
+        }
+        const userNote = note.slice(4).trim() || void 0;
+        const payload = buildReport2({
+          errorMessage: err.message,
+          httpStatus: err.httpStatus,
+          errorCode: err.code,
+          sessionId: sessionIdRef.current ?? void 0,
+          userNote,
+          model: cfg?.model,
+          cloudMode: cfg?.cloudMode
+        });
+        void sendReport(payload, cfg?.cloudToken).then((result) => {
+          setEvents((e) => [
+            ...e,
+            { kind: result.ok ? "info" : "error", key: mkKey(), text: result.message }
+          ]);
+          if (result.ok) {
+            lastApiErrorRef.current = null;
+          }
+        });
+        setEvents((e) => [
+          ...e,
+          { kind: "info", key: mkKey(), text: "Sending report\u2026" }
+        ]);
+        return true;
+      }
       if (c === "/logout") {
         unlink4(configPath()).catch(() => {
         });
@@ -19967,15 +20075,11 @@ ${lines.join("\n")}` }]);
                 { kind: "cloud_quota_exhausted", key: mkKey(), used, limit, expiresAt }
               ]);
             } else if (e instanceof KimiApiError && (e.httpStatus === 429 || e.code === 3040 || e.httpStatus !== void 0 && e.httpStatus >= 500)) {
+              const err = { httpStatus: e.httpStatus, code: e.code, message: humanizeCloudflareError(e) };
+              lastApiErrorRef.current = err;
               setEvents((es) => [
                 ...es,
-                {
-                  kind: "api_error",
-                  key: mkKey(),
-                  httpStatus: e.httpStatus,
-                  code: e.code,
-                  message: humanizeCloudflareError(e)
-                }
+                { kind: "api_error", key: mkKey(), ...err }
               ]);
             } else {
               const displayText2 = e instanceof KimiApiError ? humanizeCloudflareError(e) : e.message ?? String(e);
@@ -20393,6 +20497,7 @@ var init_app = __esm({
     init_errors();
     init_abort_scope();
     init_logger();
+    init_report2();
     init_chat();
     init_status();
     init_permission();
@@ -20671,7 +20776,7 @@ async function main() {
     cloudToken = cloudCreds.accessToken;
     cloudDeviceId = cloudCreds.deviceId;
     cfg = {
-      ...cfg ?? { accountId: "", apiToken: "", model: DEFAULT_MODEL },
+      ...cfg ?? { accountId: "", apiToken: "", model: DEFAULT_MODEL, memoryEnabled: true },
       cloudMode: true
     };
   }