mioku-plugin-chat 2.8.0 → 2.8.2

package/configs/personalization.ts

@@ -85,7 +85,7 @@ export const PERSONALIZATION_CONFIG: {
 
   replyStyle: {
     baseStyle:
-      "Casual and cute, uses emoticons, can occasionally mix in a small amount of natural everyday Japanese words, but should not heavily rely on Japanese. Do not end sentences with commas or periods.",
+      "Casual and cute,can occasionally mix in a small amount of natural everyday Japanese words, but should not heavily rely on Japanese. Do not end sentences with commas or periods.",
     multipleStyles: [
       "Playing cute, likes to add 'w' at the end of cute phrases, commonly used to replace sentence-ending particles such as '呀'.",
       "Hometown dialect mode, can occasionally use a small amount of natural everyday Japanese expressions in replies, and starts replies with '呐'. Avoid ending sentences with commas or periods.",

package/core/chat-engine.ts

@@ -1,7 +1,4 @@
-import type {
-  AIInstance,
-  SessionToolDefinition,
-} from "mioku";
+import type { AIInstance, SessionToolDefinition } from "mioku";
 import { logger } from "mioki";
 import type { AITool } from "mioku";
 import type {
@@ -11,16 +8,10 @@ import type {
   ChatResult,
 } from "../types";
 import type { HumanizeEngine } from "../humanize";
-import type {
-  StaticPromptContext,
-  DynamicPromptContext,
-} from "./prompt";
+import type { StaticPromptContext, DynamicPromptContext } from "./prompt";
 import type { SkillSessionManager } from "../manage/skill-session";
 import { createTools } from "./tools";
-import {
-  buildStaticSystemPrompt,
-  buildDynamicUserContext,
-} from "./prompt";
+import { buildStaticSystemPrompt, buildDynamicUserContext } from "./prompt";
 import type { PromptCtxForRunChat } from "../manage/types";
 import {
   isExternalSkillAllowed,
@@ -86,7 +77,6 @@ export async function runChat(
 
   const staticCtx: StaticPromptContext = {
     config: promptCtx.config,
-    botNickname: promptCtx.botNickname,
     aiService: promptCtx.aiService,
     enableExternalSkills: promptCtx.config.enableExternalSkills,
     triggerSkillRole: toolCtx.triggerSkillRole,
@@ -420,7 +410,9 @@ function estimateChatHistoryTokens(history: ChatMessage[]): number {
   );
 }
 
-function estimateMessageContentTokens(messages: Array<{ content?: unknown }>): number {
+function estimateMessageContentTokens(
+  messages: Array<{ content?: unknown }>,
+): number {
   return messages.reduce(
     (sum, message) => sum + estimateContentTokens(message.content),
     0,
@@ -474,7 +466,10 @@ function buildCurrentMessages(
   if (currentUserMessages.length > 0) {
     messages.push(
       ...prependDynamicContextToFirstUserMessage(
-        attachImagesToCurrentUserMessages(currentUserMessages, pendingImageUrls),
+        attachImagesToCurrentUserMessages(
+          currentUserMessages,
+          pendingImageUrls,
+        ),
         dynamicUserContext,
         pendingImageUrls,
       ),
@@ -651,6 +646,7 @@ function createExternalSkillRuntimeContext(toolCtx: ToolContext): any {
     rawEvent,
     session_id: toolCtx.sessionId,
     trigger_role: toolCtx.triggerSkillRole,
+    isMultimodal: Boolean(toolCtx.config?.isMultimodal),
   };
 }
 
@@ -722,9 +718,76 @@ function cleanMarkers(text: string): string {
     .replace(/<Ai>\s*<think>[\s\S]*?<\/Ai>/gi, "")
     .replace(/<｜｜DSML｜｜tool_calls>[\s\S]*?<\/｜｜DSML｜｜tool_calls>/gi, "")
     .replace(/<｜｜DSML｜｜invoke[^>]*>[\s\S]*?<\/｜｜DSML｜｜invoke>/gi, "")
-    .replace(/<｜｜DSML｜｜parameter[^>]*>[\s\S]*?<\/｜｜DSML｜｜parameter>/gi, "");
+    .replace(
+      /<｜｜DSML｜｜parameter[^>]*>[\s\S]*?<\/｜｜DSML｜｜parameter>/gi,
+      "",
+    );
+
+  return sanitizeBrackets(cleaned);
+}
+
+const FUNCTIONAL_BRACKET_PREFIX =
+  /\[(at|reply|poke|audio|emotion):[^\]\n]*\]?/gi;
 
-  return cleaned;
+function sanitizeBrackets(text: string): string {
+  if (!text) return text;
+
+  const placeholders: string[] = [];
+  let working = text.replace(FUNCTIONAL_BRACKET_PREFIX, (match) => {
+    const idx = placeholders.length;
+    placeholders.push(match);
+    return `${idx}`;
+  });
+
+  let result = "";
+  let orphanLeftCount = 0;
+  let i = 0;
+
+  while (i < working.length) {
+    const ch = working[i];
+
+    if (ch === "") {
+      const endIdx = working.indexOf("", i + 1);
+      if (endIdx > i) {
+        const idx = parseInt(working.slice(i + 1, endIdx), 10);
+        if (!Number.isNaN(idx) && placeholders[idx] !== undefined) {
+          result += placeholders[idx];
+          i = endIdx + 1;
+          continue;
+        }
+      }
+      i++;
+      continue;
+    }
+
+    if (ch === "[") {
+      let j = i + 1;
+      while (j < working.length && working[j] !== "]") {
+        j++;
+      }
+      if (j < working.length) {
+        i = j + 1;
+      } else {
+        orphanLeftCount++;
+        i++;
+      }
+      continue;
+    }
+
+    if (ch === "]") {
+      i++;
+      continue;
+    }
+
+    result += ch;
+    i++;
+  }
+
+  if (orphanLeftCount > 0) {
+    result += "]".repeat(orphanLeftCount);
+  }
+
+  return result;
 }
 
 function removeStickerIntentLines(text: string): string {

package/core/prompt.ts

@@ -15,13 +15,12 @@ import type { SkillSessionManager } from "../manage/skill-session";
  * Everything here is required to be identical across consecutive requests for the same bot
  * so OpenAI's auto prompt caching can hit the system block.
  *
- * Concretely: persona, config flags, enabled features, allowed external skills, persona style.
- * Anything that changes turn-to-turn (time, group, history, target, emotion, replies context, etc.)
+ * Concretely: persona, config flags, enabled features, allowed external skills.
+ * Anything that changes turn-to-turn (time, group, history, target, emotion, reply style, replies context, etc.)
  * belongs in DynamicPromptContext instead.
  */
 export interface StaticPromptContext {
   config: ChatConfig;
-  botNickname: string;
   aiService: AIService;
   enableExternalSkills: boolean;
   triggerSkillRole?: SkillPermissionRole;
@@ -144,17 +143,16 @@ const REPLY_STYLE_LENGTH: Record<Strength, string> = {
 };
 
 const AUDIO_MODE_LINE: Record<Strength, string> = {
-  high:
-    "- Use voice sparingly. Only use it when spoken delivery is clearly better than text, such as a greeting, a sharp emotional reaction, or a daily phrase.",
+  high: "- Use voice sparingly. Only use it when spoken delivery is clearly better than text, such as a greeting, a sharp emotional reaction, or a daily phrase.",
   medium:
     "- You may use voice for greetings, reactions, calls, confirmations, or comforting words, but stay selective.",
   low: "- When a short spoken reaction would make the conversation feel more natural or vivid, you can use voice more freely.",
 };
 
 const MARKDOWN_MODE_LINE: Record<Strength, string> = {
-  high:
-    "- Prefer normal chat text. Use Markdown only when the reply truly needs structured presentation, such as a tutorial, comparison, detailed explanation, code sample or processing large amounts of data, such as after a web search or viewing a webpage.",
-  medium: "- Use Markdown when your responses require a structured presentation.",
+  high: "- Prefer normal chat text. Use Markdown only when the reply truly needs structured presentation, such as a tutorial, comparison, detailed explanation, code sample or processing large amounts of data, such as after a web search or viewing a webpage.",
+  medium:
+    "- Use Markdown when your responses require a structured presentation.",
   low: "- Use Markdown freely where it can make your responses clearer.",
 };
 
@@ -175,15 +173,13 @@ const TOOL_INTENSITY_BLOCK: Record<Strength, string> = {
 };
 
 const EMOJI_MODE_LINE: Record<Strength, string> = {
-  high:
-    "- Do not use stickers in consecutive chat turns. Use a sticker only when you are in an emotionally intense state.",
+  high: "- Do not use stickers in consecutive chat turns. Use a sticker only when you are in an emotionally intense state.",
   medium: "- Do not use stickers in consecutive chat turns.",
   low: "- Feel free to use a sticker whenever you want.",
 };
 
 const WEB_SEARCH_LINE: Record<Strength, string> = {
-  high:
-    "- When facts may be outdated or uncertain, proactively call web_search instead of guessing.",
+  high: "- When facts may be outdated or uncertain, proactively call web_search instead of guessing.",
   medium: "- Use web_search when current or external info is needed.",
   low: "- Use web_search only when the user explicitly needs external/current information.",
 };
@@ -210,7 +206,6 @@ export function buildStaticSystemPrompt(ctx: StaticPromptContext): string {
   const sections: string[] = [];
 
   sections.push(buildPersonaSection(ctx.config.persona));
-  sections.push(buildReplyStyleSection(ctx.config, ctx.botNickname, lengthStrength));
   sections.push(
     buildResponseFormatSection(
       ctx,
@@ -244,7 +239,7 @@ function buildReplyStyleSection(
  ### Behavior Guidelines
  - You are a participant in this chat, not an assistant. Chat naturally like a real person.
  - ${REPLY_STYLE_LENGTH[lengthStrength]}
- - Match the language used by others in the chat (if they speak Chinese, reply in Chinese, etc.).
+ - Match the language used by others in the chat.
  - Don't repeat yourself or echo what others just said.
  - **NEVER use action descriptions like *xxx* or (xxx) — just speak as a normal person would**
  - **${markdownBehaviorLine(config)}**
@@ -272,31 +267,24 @@ function buildResponseFormatSection(
 
   lines.push(`Your text response IS your reply to the chat. It will be sent directly as a message.
 - **IMPORTANT: Output ONLY your final reply text. Do NOT include your thinking process, reasoning, analysis, or internal thoughts.**
-- Do NOT prefix your response with phrases like "Let me think", "I should", "I need to", "Based on", "Looking at", etc.
 - Do NOT explain what you're doing or why. Just say what you want to say directly.
-- **MULTIPLE MESSAGES (CRITICAL!): Each line (separated by Enter/Return) will be sent as a SEPARATE message.**
+- **MULTIPLE MESSAGES: Each line (separated by Enter/Return) will be sent as a SEPARATE message.**
   - If you want to send multiple messages, just press Enter and write the next line
-  - Each line = one message sent to the chat
   - **If your reply has multiple sentences or different points, ALWAYS use real line breaks to separate them**
-  - NEVER use "\\" or literal "\\n" to simulate a new line
-- **MESSAGE ORDER MATTERS**: messages are sent top-to-bottom, one line at a time.
-- For action markers like [] or [audio:...], put them on their own line when they are meant to be a separate action.
+- For action markers like [] , put them on their own line when they are meant to be a separate action.
 
 - **SPECIAL ACTIONS in your text (auto-parsed and removed from message):**
   - Use [at:123456] in your text to @ someone (123456 is the QQ number)
-  - Use [poke:123456] in your text to poke someone. IMPORTANT: when you plan to poke a user, DON't emphasize words like "戳你一下 or 戳回去" to describe your actions
+  - Use [poke:123456] in your text to poke someone. IMPORTANT: when you plan to poke a user, DON't describe your poke actions.
   - Use [reply:123456] at the START of a line to quote-reply that message (123456 is message_id)
-  - **You can use MULTIPLE [reply:xxx] markers in different lines to quote multiple messages!**
-  - These markers will be automatically parsed and removed from your sent message`);
+  - **You can use MULTIPLE [reply:xxx] markers in different lines to quote multiple messages!**`);
 
   if (ctx.config.audio?.enabled && ctx.config.audio.baseUrl?.trim()) {
     lines.push(`
 ### Optional Voice Message Format
 - You MAY optionally send one voice message by writing [audio:content]
 - Audio is OPTIONAL. Do NOT use it in every reply
-The voice message function sends plain text and cannot be used for singing. If a user needs you to sing, other skills should be considered first.
-- Put [audio:...] on its own line when you want it sent as a separate message in sequence
-- Example: "[audio:おはようー]"
+The voice message function sends plain text and cannot be used for singing。
 ${AUDIO_MODE_LINE[audioStrength]}`);
   }
 
@@ -316,9 +304,6 @@ ${MARKDOWN_MODE_LINE[markdownStrength]}
   lines.push(`
 ### Tool Calling Format
 - When you decide to use a tool, you MUST use the structured tool_calls mechanism provided by the API
-- Do NOT output tool calls, tool names, or tool arguments in your reply text under any circumstances
-- Do NOT use XML, JSON, or any text format to describe tool calls — only use the API's tool_calls field
-- Each tool's description contains its own usage guidance; read those before calling a tool. If a tool's description says "use only when X" or "do not call for every question", respect that.
 - web_search and web_read_page are limited per conversation; do not retry excessively`);
 
   appendEmojiSection(lines, ctx, emojiStrength);
@@ -343,9 +328,7 @@ function appendEmojiSection(
 ### Optional Sticker / Emoji Format
 - You MAY optionally request one matching sticker by writing exactly [] on its own line
 ${EMOJI_MODE_LINE[emojiStrength]}
-- Never put an emotion, label, character, or any other text inside the brackets
-- Output at most one [] block in a turn
-- A separate sticker selection agent will read the current conversation and choose the actual sticker`);
+- Never put an emotion, label, character, or any other text inside the brackets`);
 }
 
 function appendExternalSkillsSection(
@@ -455,6 +438,9 @@ export function buildDynamicUserContext(ctx: DynamicPromptContext): string {
     sections.push(`## Planner's Analysis\n${ctx.plannerThoughts}`);
   }
 
+  sections.push(
+    buildReplyStyleSection(ctx.config, ctx.botNickname, lengthStrength),
+  );
   sections.push(buildEmotionSection(ctx));
 
   return sections.join("\n\n");
@@ -540,7 +526,6 @@ function buildPokedGuidance(length: Strength): string[] {
   return [
     "Someone pokes you in a group, probably out of non-malicious play or to draw your attention to what happened in the group chat.",
     "Don't make a fuss about replying, just observe whether the chat history in the group has noteworthy content, and if not, simply say hello or express concern to the user.",
-    "Reply naturally in combination with the context, don't say something like \"怎么又来戳我了\"",
     POKED_LENGTH[length],
   ].filter(Boolean);
 }
@@ -622,13 +607,8 @@ function buildChatHistorySection(ctx: DynamicPromptContext): string {
   }
   flushAssistant();
 
-  return `## Recent Context (Only reference if directly relevant)
-Just the last few messages - don't overthink it or dig into old conversations:
-
+  return `## Recent Context
 ${mergedLines.join("\n")}
-
-Note: Messages may contain media tags like [image:描述], [video:描述], [forward:摘要], [card:摘要], or [group_notice:摘要]. If you need detailed information about an image or video, use the view_media tool with the message ID.
-
 -- DON'T repeat yourself or bring up old topics - focus on what's being said right now. --`;
 }
 
@@ -657,7 +637,7 @@ ${messageBlocks.join("\n")}
 IMPORTANT: You do NOT need to reply to each person or each message above. Give ONE casual response to the group as a whole.`;
   }
 
-  return `## >>> Target Message (Reply to THIS) <<<
+  return `## >>> Target Message <<<
 [${timeStr}] ${target.userName}(${target.userId}, ${target.userRole}${target.userTitle ? `, ${target.userTitle}` : ""})${msgIdStr}: ${target.content}`;
 }
 
@@ -687,7 +667,7 @@ function buildEmotionSection(ctx: DynamicPromptContext): string {
     lines.push(`Available emotions: ${availableEmotions.join(", ")}`);
   }
   lines.push(
-    "You may switch your emotion state by writing [emotion:emotion_name]. The marker is not sent to the chat. Only use available emotions.",
+    "You may switch your emotion state by writing [emotion:emotion_name].",
   );
 
   if (examples.length > 0) {
@@ -702,7 +682,9 @@ function buildEmotionSection(ctx: DynamicPromptContext): string {
 }
 
 function normalizeEmotionName(value: unknown): string {
-  return String(value || "").trim().toLowerCase();
+  return String(value || "")
+    .trim()
+    .toLowerCase();
 }
 
 function normalizeEmotionExamples(value: unknown): string[] {
@@ -744,7 +726,5 @@ export function buildRecallMemoryFeatureSection(config: ChatConfig): string {
   return `
 ### Memory Recall Tool
 - recall_memory: Delegate recall to a memory worker model. Pass a clear recall question and let the worker search historical logs.
-- Use recall_memory ONLY when there is explicit need to recall past content and required information is clearly missing from current context.
-- Do NOT call recall_memory for every question.
-- The worker returns historical logs with timestamps; treat them as past records, not newly sent messages.`;
+- Use recall_memory ONLY when there is explicit need to recall past content and required information is clearly missing from current context.`;
 }

package/core/tools/info.ts

@@ -161,11 +161,11 @@ export function createInfoTools(toolCtx: ToolContext): AITool[] {
             );
           }
 
-          const ai = toolCtx.aiService.getDefault();
-          if (!ai) return { error: "AI instance not available" };
+          const visionAI = toolCtx.aiService.getInstanceByRole?.("vision") ?? toolCtx.aiService.getDefault();
+          if (!visionAI) return { error: "AI instance not available" };
 
           const result = await describeImage(
-            ai,
+            visionAI,
             media.url,
             toolCtx.config.multimodalWorkingModel,
             toolCtx.event?.raw_message || undefined,
@@ -214,11 +214,11 @@ export function createInfoTools(toolCtx: ToolContext): AITool[] {
             );
           }
 
-          const ai = toolCtx.aiService.getDefault();
-          if (!ai) return { error: "AI instance not available" };
+          const visionAI = toolCtx.aiService.getInstanceByRole?.("vision") ?? toolCtx.aiService.getDefault();
+          if (!visionAI) return { error: "AI instance not available" };
 
           const summary = await summarizeVideoContent(videoFile.path, videoFile.byteSize, {
-            ai,
+            ai: visionAI,
             multimodalWorkingModel: toolCtx.config.multimodalWorkingModel,
             logger: {
               info: (m) => logger.info(m),
@@ -269,11 +269,11 @@ export function createInfoTools(toolCtx: ToolContext): AITool[] {
         }
 
         const { describeImage } = await import("../multimodal");
-        const ai = toolCtx.aiService.getDefault();
-        if (!ai) return { error: "AI instance not available" };
+        const visionAI = toolCtx.aiService.getInstanceByRole?.("vision") ?? toolCtx.aiService.getDefault();
+        if (!visionAI) return { error: "AI instance not available" };
 
         const result = await describeImage(
-          ai,
+          visionAI,
           avatarUrl,
           toolCtx.config.multimodalWorkingModel,
           `User ${args.user_id}'s QQ avatar`,

package/core/tools/web.ts

@@ -234,7 +234,7 @@ export function createWebReadPageTool(toolCtx: ToolContext): AITool {
     handler: async (args) => {
       try {
         const ai = toolCtx.config.webReader.useWorkingModel
-          ? toolCtx.aiService.getDefault()
+          ? (toolCtx.aiService.getInstanceByRole?.("working") ?? toolCtx.aiService.getDefault())
           : undefined;
         if (toolCtx.config.webReader.useWorkingModel && !ai) {
           return { success: false, error: "AI instance not available" };
@@ -275,7 +275,7 @@ export function createRecallMemoryTool(toolCtx: ToolContext): AITool {
       const question = String(args?.question || "").trim();
       if (!question) return { success: false, error: "question is required" };
 
-      const ai = toolCtx.aiService.getDefault();
+      const ai = toolCtx.aiService.getInstanceByRole?.("working") ?? toolCtx.aiService.getDefault();
       if (!ai) return { success: false, error: "AI instance not available" };
 
       const groupHistoryLimit = resolveGroupRecallLimit(toolCtx.config.memory?.groupHistoryLimit);

package/handlers/message.ts

@@ -68,11 +68,11 @@ export function createMessageHandler(
 
     // 媒体分析
     if (isGroup && groupId && e.message && !isMediaAnalysisBlocked(cfg, userId)) {
-      const ai = pluginCtx.aiService.getDefault();
+      const visionAI = pluginCtx.visionAIInstance ?? pluginCtx.aiService.getDefault();
       const bot = ctx.pickBot(e.self_id) as any;
-      const mediaOptions = ai
+      const mediaOptions = visionAI
         ? buildHistoryMediaProcessingOptions(
-            ai,
+            visionAI,
             cfg,
             pluginCtx.db,
             bot,
@@ -91,13 +91,13 @@ export function createMessageHandler(
           )
         : undefined;
 
-      if (ai && cfg.enableMediaRecognition) {
+      if (visionAI && cfg.enableMediaRecognition) {
         const { processImage } = await import("../core/media/image-analyzer");
         for (const seg of e.message) {
           if (seg.type === "image") {
             const imageUrl = getSegmentUrl(seg);
             if (imageUrl) {
-              processImage(ai, imageUrl, cfg.multimodalWorkingModel, pluginCtx.db, {
+              processImage(visionAI, imageUrl, cfg.multimodalWorkingModel, pluginCtx.db, {
                 runAIRequest: (request) =>
                   pluginCtx.runWithRateLimitGuard(request, {
                     userId,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mioku-plugin-chat",
-  "version": "2.8.0",
+  "version": "2.8.2",
   "description": "AI 智能聊天插件",
   "main": "index.ts",
   "type": "module",