chore: release v2.5.2

Stop Codex from satisfying a request by copying an unrelated pre-existing
image from generated_images instead of generating a new one. Verification
now requires a PNG in this thread's generated_images dir (or an image_gen
stream item, kept as a forward-compatible signal), and the spawned-agent
instruction forbids reading or reusing history images before image_gen is
called. Removes the findCpToTarget fallback that allowed the shortcut.

Validated against a real codex exec run: image_gen leaves no stream item,
so the filesystem check is the load-bearing signal; locked in via a
condensed real-stream regression fixture.

Closes #185

.claude-plugin/marketplace.json

@@ -6,7 +6,7 @@
   },
   "metadata": {
     "description": "Skills shared by Baoyu for improving daily work efficiency",
-    "version": "2.5.1"
+    "version": "2.5.2"
   },
   "plugins": [
     {

CHANGELOG.md

@@ -2,6 +2,14 @@
 
 English | [中文](./CHANGELOG.zh.md)
 
+## 2.5.2 - 2026-06-18
+
+### Fixes
+- `codex-imagegen`: stop Codex from satisfying a request by copying an unrelated pre-existing image from `generated_images` instead of generating a new one. Verification now requires real evidence that `image_gen` ran in the current thread — an `image_gen` event in the stream **or** a PNG in this thread's `generated_images` dir — and the spawned-agent instruction forbids reading or reusing history images before `image_gen` is called ([#185](https://github.com/JimLiu/baoyu-skills/issues/185))
+
+### Documentation
+- README: document Codex project-level install (`.agents/skills`) and project- vs user-level WeChat credential scopes
+
 ## 2.5.1 - 2026-06-13
 
 ### Documentation

CHANGELOG.zh.md

@@ -2,6 +2,14 @@
 
 [English](./CHANGELOG.md) | 中文
 
+## 2.5.2 - 2026-06-18
+
+### 修复
+- `codex-imagegen`：阻止 Codex 通过复制 `generated_images` 中无关旧图片来代替真正生成的取巧行为。校验改为要求当前线程内 `image_gen` 真正运行的证据——事件流中的 `image_gen` 事件，**或**当前线程 `generated_images` 目录下新生成的 PNG——并在派生 agent 指令中禁止在调用 `image_gen` 之前读取或复用历史图片（[#185](https://github.com/JimLiu/baoyu-skills/issues/185)）
+
+### 文档
+- README：补充 Codex 项目级安装（`.agents/skills`）以及项目级与全局公众号凭证的放置位置
+
 ## 2.5.1 - 2026-06-13
 
 ### 文档

README.md

@@ -19,6 +19,31 @@ Skills shared by Baoyu for improving daily work efficiency with AI Agents (Claud
 npx skills add jimliu/baoyu-skills
 ```
 
+### Codex Project-Level Install
+
+If you only need a subset of skills in one project, you do not need to install the full plugin. Codex scans `.agents/skills` inside a project, so copy or symlink each needed skill as a full directory:
+
+```text
+<project>/.agents/skills/baoyu-cover-image/SKILL.md
+<project>/.agents/skills/baoyu-article-illustrator/SKILL.md
+<project>/.agents/skills/baoyu-post-to-wechat/SKILL.md
+```
+
+For a WeChat Official Account article workflow, the usual minimal set is:
+
+- `baoyu-cover-image`
+- `baoyu-article-illustrator`
+- `baoyu-post-to-wechat`
+
+You do not need to install `baoyu-markdown-to-html` separately. `baoyu-post-to-wechat` already includes the Markdown to WeChat-ready HTML conversion flow. Install `baoyu-format-markdown` only if you need to first turn raw text or drafts into structured Markdown articles with titles, summaries, headings, bold text, lists, and similar formatting.
+
+Place WeChat API credentials according to the scope you want:
+
+- User-level: `~/.baoyu-skills/.env`
+- Project-level: `<project>/.baoyu-skills/.env`
+
+Project-level `.env` files are useful when credentials should apply only to the current project. Do not commit them to Git.
+
 ### Publish to ClawHub / OpenClaw
 
 This repository now supports publishing each `skills/baoyu-*` directory as an individual ClawHub skill.

README.zh.md

@@ -29,6 +29,31 @@
 npx skills add jimliu/baoyu-skills
 ```
 
+### Codex 项目级安装
+
+如果只在某个项目中使用部分技能，不需要安装整个插件。Codex 会扫描项目里的 `.agents/skills`，可以只把需要的 skill 按整个目录复制或软链接到当前项目：
+
+```text
+<project>/.agents/skills/baoyu-cover-image/SKILL.md
+<project>/.agents/skills/baoyu-article-illustrator/SKILL.md
+<project>/.agents/skills/baoyu-post-to-wechat/SKILL.md
+```
+
+公众号文章发布的最小组合通常是：
+
+- `baoyu-cover-image`
+- `baoyu-article-illustrator`
+- `baoyu-post-to-wechat`
+
+不需要单独安装 `baoyu-markdown-to-html`，`baoyu-post-to-wechat` 已内置 Markdown 到公众号可用 HTML 的转换流程。只有需要先把原始文本或草稿整理成 Markdown 文章结构（标题、摘要、层级标题、加粗、列表等）时，再额外安装 `baoyu-format-markdown`。
+
+公众号 API 凭证按作用范围放置：
+
+- 全局：`~/.baoyu-skills/.env`
+- 项目：`<project>/.baoyu-skills/.env`
+
+项目级 `.env` 适合只给当前项目使用，注意不要提交到 Git。
+
 ### 发布到 ClawHub / OpenClaw
 
 现在这个仓库支持把每个 `skills/baoyu-*` 目录作为独立 ClawHub skill 发布。

packages/baoyu-codex-imagegen/src/main.ts

@@ -6,7 +6,7 @@ import process from "node:process";
 import { setTimeout as delay } from "node:timers/promises";
 import { GenError, type CliOptions, type GenerateResult } from "./types.ts";
 import { runCodexExec } from "./spawn.ts";
-import { findCpToTarget, verifyImageGenWasInvoked, verifyOutput } from "./validator.ts";
+import { hasImageGenEvidence, verifyImageGenWasInvoked, verifyOutput } from "./validator.ts";
 import { cacheKey, lookupCache, storeCache, FileLock } from "./cache.ts";
 import { JsonLogger } from "./logger.ts";
 
@@ -125,7 +125,7 @@ function buildInstruction(prompt: string, opts: CliOptions): string {
   const refHint = opts.refImages.length > 0
     ? `\nREFERENCE IMAGES (attached above): ${opts.refImages.length} image(s) provided for style/composition guidance.\n`
     : "";
-  return `You have an internal tool called image_gen for image generation. Use it.
+  return `You have an internal tool called image_gen for image generation. You MUST call it before doing anything else.
 
 TASK: Generate an image with the spec below, then save to disk.
 
@@ -137,12 +137,15 @@ OUTPUT PATH: ${opts.outputPath}
 ${refHint}
 STEPS:
 1. Call image_gen with the prompt and aspect ratio above${opts.refImages.length > 0 ? " (using the attached reference images for guidance)" : ""}.
-2. Move or copy the resulting image from Codex default location ($CODEX_HOME/generated_images/...) to: ${opts.outputPath}
+2. Move or copy ONLY the image produced by that image_gen call from Codex default location ($CODEX_HOME/generated_images/...) to: ${opts.outputPath}
 3. Verify with: ls -la ${opts.outputPath}
 4. Reply with ONLY this JSON line (no markdown fences, no other text):
    {"status":"ok","path":"${opts.outputPath}","bytes":<file_size_in_bytes>}
 
 HARD CONSTRAINTS:
+- Do NOT search for, find, inspect, reuse, or copy any pre-existing files from $CODEX_HOME/generated_images/ or any other directory.
+- Do NOT run ls/find/rg/grep/glob over $CODEX_HOME/generated_images/ before image_gen has been called.
+- You MUST call image_gen first. Only after image_gen completes may you copy the newly created file from this turn.
 - Do NOT use curl, wget, Python, or any external API.
 - Do NOT use bash to fabricate an image; only image_gen produces real pixels.
 - Use ONLY the image_gen internal tool.`;
@@ -175,13 +178,16 @@ async function attemptGenerate(
     throw new GenError("agent_refused", "No thread id in event stream");
   }
 
-  // verify: image_gen was actually invoked (check $CODEX_HOME/generated_images/{threadId}/)
+  // verify image_gen ran in THIS thread. A PNG in this thread's
+  // generated_images dir is the real signal (image_gen does not surface as a
+  // stream item); the stream check is a forward-compatible fallback. The #185
+  // shortcut (copying an unrelated history image) yields neither.
   const ver = await verifyImageGenWasInvoked(run.threadId);
-  if (!ver.ok) {
+  if (!hasImageGenEvidence(run.toolCalls, ver.ok)) {
-    // secondary verify: did tool_calls include cp/mv from generated_images to our target
+    throw new GenError(
-    if (!findCpToTarget(run.toolCalls, opts.outputPath)) {
+      "no_image_gen_tool_use",
-      throw new GenError("no_image_gen_tool_use", `image_gen was not invoked: ${ver.reason}`);
+      `image_gen was not invoked (no image_gen event in stream; ${ver.reason})`,
-    }
+    );
   }
 
   // verify output

packages/baoyu-codex-imagegen/src/parser.test.ts

@@ -33,17 +33,26 @@ test("parseEventStream tolerates malformed lines", () => {
   expect(r.usage?.input).toBe(1);
 });
 
-test("hasImageGenInvocation finds shell calls touching generated_images", () => {
+test("hasImageGenInvocation does not infer image_gen from shell copies", () => {
   const r = parseEventStream(REAL_PoC_STREAM);
-  // image_gen itself is not an event; inferred via generated_images cp path
-  // this test only verifies parser behavior; driver logic lives in validator
   const hasCp = r.toolCalls.some((tc) => tc.command?.includes("generated_images"));
   expect(hasCp).toBe(true);
+  expect(hasImageGenInvocation(r.toolCalls)).toBe(false);
 });
 
-test("hasImageGenInvocation (proper) returns false when no image_gen tool", () => {
+test("hasImageGenInvocation detects real image_gen tool calls only", () => {
   expect(hasImageGenInvocation([{ id: "1", tool: "shell", status: "completed" }])).toBe(false);
   expect(
     hasImageGenInvocation([{ id: "1", tool: "image_gen", status: "completed" }]),
   ).toBe(true);
 });
+
+test("parseEventStream normalizes an image_generation item to image_gen", () => {
+  const stream = `{"type":"thread.started","thread_id":"t1"}
+{"type":"item.started","item":{"id":"g0","type":"image_generation","status":"in_progress"}}
+{"type":"item.completed","item":{"id":"g0","type":"image_generation","status":"completed"}}
+{"type":"item.completed","item":{"id":"m0","type":"agent_message","text":"done"}}`;
+  const r = parseEventStream(stream);
+  expect(r.toolCalls.some((tc) => tc.tool === "image_gen")).toBe(true);
+  expect(hasImageGenInvocation(r.toolCalls)).toBe(true);
+});

packages/baoyu-codex-imagegen/src/validator.test.ts

@@ -2,11 +2,52 @@ import { test, expect } from "bun:test";
 import { mkdtemp, writeFile, rm, mkdir } from "node:fs/promises";
 import { tmpdir } from "node:os";
 import path from "node:path";
-import { verifyOutput, verifyImageGenWasInvoked, findCpToTarget } from "./validator.ts";
+import { verifyOutput, verifyImageGenWasInvoked, hasImageGenEvidence } from "./validator.ts";
+import { parseEventStream } from "./parser.ts";
 import { GenError } from "./types.ts";
 
 const PNG_HEADER = Buffer.from([0x89, 0x50, 0x4e, 0x47, 0x0d, 0x0a, 0x1a, 0x0a]);
 
+// Condensed + sanitized capture of a REAL successful `codex exec --json` run
+// (thread 019edc1c…, 16:9 maple-tree image, single attempt). The image_gen
+// tool leaves NO stream item — success shows only reasoning / command_execution
+// / agent_message, and a `cp` from generated_images/<FULL-thread-id>/ig_*.png.
+const REAL_SUCCESS_STREAM = `{"type":"thread.started","thread_id":"019edc1c-e7a3-74f0-a276-13bea71d32d6"}
+{"type":"turn.started"}
+{"type":"item.completed","item":{"id":"r0","type":"reasoning"}}
+{"type":"item.completed","item":{"id":"r1","type":"reasoning"}}
+{"type":"item.completed","item":{"id":"m0","type":"agent_message","text":"Image generation is complete. Locating the newly produced image and copying it to the requested path."}}
+{"type":"item.started","item":{"id":"c0","type":"command_execution","command":"ls $CODEX_HOME/generated_images","status":"in_progress"}}
+{"type":"item.completed","item":{"id":"c0","type":"command_execution","command":"ls $CODEX_HOME/generated_images","exit_code":1,"status":"failed"}}
+{"type":"item.started","item":{"id":"c1","type":"command_execution","command":"cp $CODEX_HOME/generated_images/019edc1c-e7a3-74f0-a276-13bea71d32d6/ig_03eda661.png /Users/x/out/maple.png","status":"in_progress"}}
+{"type":"item.completed","item":{"id":"c1","type":"command_execution","command":"cp $CODEX_HOME/generated_images/019edc1c-e7a3-74f0-a276-13bea71d32d6/ig_03eda661.png /Users/x/out/maple.png","exit_code":0,"status":"completed"}}
+{"type":"item.completed","item":{"id":"m1","type":"agent_message","text":"{\\"status\\":\\"ok\\",\\"path\\":\\"/Users/x/out/maple.png\\",\\"bytes\\":1317377}"}}
+{"type":"turn.completed","usage":{"input_tokens":49489,"cached_input_tokens":34432,"output_tokens":2463,"reasoning_output_tokens":1990}}`;
+
+test("real success stream carries no image_gen item — gating on the stream alone would false-negative (#185)", () => {
+  const r = parseEventStream(REAL_SUCCESS_STREAM);
+  expect(r.threadId).toBe("019edc1c-e7a3-74f0-a276-13bea71d32d6");
+  // success path tools: reasoning / shell / agent_message — never image_gen
+  expect(r.toolCalls.map((tc) => tc.tool).sort()).toEqual([
+    "agent_message",
+    "agent_message",
+    "reasoning",
+    "reasoning",
+    "shell",
+    "shell",
+  ]);
+  expect(r.toolCalls.some((tc) => tc.tool === "image_gen")).toBe(false);
+  // filesystem evidence (PNG in this thread's dir) is what must let it through
+  expect(hasImageGenEvidence(r.toolCalls, true)).toBe(true);
+  // with no filesystem evidence and no stream item, it is rejected (the shortcut)
+  expect(hasImageGenEvidence(r.toolCalls, false)).toBe(false);
+});
+
+test("hasImageGenEvidence accepts a real image_gen stream item even without a dir PNG", () => {
+  expect(hasImageGenEvidence([{ id: "1", tool: "image_gen", status: "completed" }], false)).toBe(true);
+  expect(hasImageGenEvidence([{ id: "1", tool: "shell", status: "completed" }], false)).toBe(false);
+});
+
 test("verifyOutput passes for valid PNG", async () => {
   const dir = await mkdtemp(path.join(tmpdir(), "cig-val-"));
   try {
@@ -73,26 +114,3 @@ test("verifyImageGenWasInvoked true when PNG exists in thread dir", async () =>
     await rm(tempHome, { recursive: true, force: true });
   }
 });
-
-test("findCpToTarget detects cp from generated_images", () => {
-  expect(
-    findCpToTarget(
-      [
-        {
-          id: "1",
-          tool: "shell",
-          status: "completed",
-          command: "cp ~/.codex/generated_images/thread/ig_x.png /tmp/out.png",
-        },
-      ],
-      "/tmp/out.png",
-    ),
-  ).toBe(true);
-
-  expect(
-    findCpToTarget(
-      [{ id: "1", tool: "shell", status: "completed", command: "ls /tmp" }],
-      "/tmp/out.png",
-    ),
-  ).toBe(false);
-});

packages/baoyu-codex-imagegen/src/validator.ts

@@ -3,6 +3,7 @@ import { homedir } from "node:os";
 import path from "node:path";
 import { GenError } from "./types.ts";
 import type { ToolCall } from "./types.ts";
+import { hasImageGenInvocation } from "./parser.ts";
 
 const PNG_MAGIC = Buffer.from([0x89, 0x50, 0x4e, 0x47, 0x0d, 0x0a, 0x1a, 0x0a]);
 
@@ -23,15 +24,15 @@ export async function verifyImageGenWasInvoked(threadId: string | null): Promise
   }
 }
 
-export function findCpToTarget(toolCalls: ToolCall[], target: string): boolean {
+// Real evidence that image_gen ran in THIS thread. Codex's image_gen tool does
-  return toolCalls.some(
+// not surface as a stream item, so a successful run shows only reasoning/shell/
-    (tc) =>
+// agent_message — `dirHasImage` (a PNG in this thread's generated_images dir) is
-      tc.tool === "shell" &&
+// what proves it. The stream check is kept as a forward-compatible signal in
-      typeof tc.command === "string" &&
+// case a future Codex version emits the item. The #185 shortcut (copying an
-      (tc.command.includes(target) || tc.command.includes(path.basename(target))) &&
+// unrelated history image, which lives under a different thread id) yields
-      /\b(cp|mv|cat)\b/.test(tc.command) &&
+// neither, so it is correctly rejected.
-      tc.command.includes("generated_images"),
+export function hasImageGenEvidence(toolCalls: ToolCall[], dirHasImage: boolean): boolean {
-  );
+  return dirHasImage || hasImageGenInvocation(toolCalls);
 }
 
 export async function verifyOutput(outputPath: string): Promise<{ bytes: number }> {

skills/baoyu-image-gen/scripts/codex-imagegen/main.ts

@@ -6,7 +6,7 @@ import process from "node:process";
 import { setTimeout as delay } from "node:timers/promises";
 import { GenError, type CliOptions, type GenerateResult } from "./types.ts";
 import { runCodexExec } from "./spawn.ts";
-import { findCpToTarget, verifyImageGenWasInvoked, verifyOutput } from "./validator.ts";
+import { hasImageGenEvidence, verifyImageGenWasInvoked, verifyOutput } from "./validator.ts";
 import { cacheKey, lookupCache, storeCache, FileLock } from "./cache.ts";
 import { JsonLogger } from "./logger.ts";
 
@@ -125,7 +125,7 @@ function buildInstruction(prompt: string, opts: CliOptions): string {
   const refHint = opts.refImages.length > 0
     ? `\nREFERENCE IMAGES (attached above): ${opts.refImages.length} image(s) provided for style/composition guidance.\n`
     : "";
-  return `You have an internal tool called image_gen for image generation. Use it.
+  return `You have an internal tool called image_gen for image generation. You MUST call it before doing anything else.
 
 TASK: Generate an image with the spec below, then save to disk.
 
@@ -137,12 +137,15 @@ OUTPUT PATH: ${opts.outputPath}
 ${refHint}
 STEPS:
 1. Call image_gen with the prompt and aspect ratio above${opts.refImages.length > 0 ? " (using the attached reference images for guidance)" : ""}.
-2. Move or copy the resulting image from Codex default location ($CODEX_HOME/generated_images/...) to: ${opts.outputPath}
+2. Move or copy ONLY the image produced by that image_gen call from Codex default location ($CODEX_HOME/generated_images/...) to: ${opts.outputPath}
 3. Verify with: ls -la ${opts.outputPath}
 4. Reply with ONLY this JSON line (no markdown fences, no other text):
    {"status":"ok","path":"${opts.outputPath}","bytes":<file_size_in_bytes>}
 
 HARD CONSTRAINTS:
+- Do NOT search for, find, inspect, reuse, or copy any pre-existing files from $CODEX_HOME/generated_images/ or any other directory.
+- Do NOT run ls/find/rg/grep/glob over $CODEX_HOME/generated_images/ before image_gen has been called.
+- You MUST call image_gen first. Only after image_gen completes may you copy the newly created file from this turn.
 - Do NOT use curl, wget, Python, or any external API.
 - Do NOT use bash to fabricate an image; only image_gen produces real pixels.
 - Use ONLY the image_gen internal tool.`;
@@ -175,13 +178,16 @@ async function attemptGenerate(
     throw new GenError("agent_refused", "No thread id in event stream");
   }
 
-  // verify: image_gen was actually invoked (check $CODEX_HOME/generated_images/{threadId}/)
+  // verify image_gen ran in THIS thread. A PNG in this thread's
+  // generated_images dir is the real signal (image_gen does not surface as a
+  // stream item); the stream check is a forward-compatible fallback. The #185
+  // shortcut (copying an unrelated history image) yields neither.
   const ver = await verifyImageGenWasInvoked(run.threadId);
-  if (!ver.ok) {
+  if (!hasImageGenEvidence(run.toolCalls, ver.ok)) {
-    // secondary verify: did tool_calls include cp/mv from generated_images to our target
+    throw new GenError(
-    if (!findCpToTarget(run.toolCalls, opts.outputPath)) {
+      "no_image_gen_tool_use",
-      throw new GenError("no_image_gen_tool_use", `image_gen was not invoked: ${ver.reason}`);
+      `image_gen was not invoked (no image_gen event in stream; ${ver.reason})`,
-    }
+    );
   }
 
   // verify output

skills/baoyu-image-gen/scripts/codex-imagegen/validator.ts

@@ -3,6 +3,7 @@ import { homedir } from "node:os";
 import path from "node:path";
 import { GenError } from "./types.ts";
 import type { ToolCall } from "./types.ts";
+import { hasImageGenInvocation } from "./parser.ts";
 
 const PNG_MAGIC = Buffer.from([0x89, 0x50, 0x4e, 0x47, 0x0d, 0x0a, 0x1a, 0x0a]);
 
@@ -23,15 +24,15 @@ export async function verifyImageGenWasInvoked(threadId: string | null): Promise
   }
 }
 
-export function findCpToTarget(toolCalls: ToolCall[], target: string): boolean {
+// Real evidence that image_gen ran in THIS thread. Codex's image_gen tool does
-  return toolCalls.some(
+// not surface as a stream item, so a successful run shows only reasoning/shell/
-    (tc) =>
+// agent_message — `dirHasImage` (a PNG in this thread's generated_images dir) is
-      tc.tool === "shell" &&
+// what proves it. The stream check is kept as a forward-compatible signal in
-      typeof tc.command === "string" &&
+// case a future Codex version emits the item. The #185 shortcut (copying an
-      (tc.command.includes(target) || tc.command.includes(path.basename(target))) &&
+// unrelated history image, which lives under a different thread id) yields
-      /\b(cp|mv|cat)\b/.test(tc.command) &&
+// neither, so it is correctly rejected.
-      tc.command.includes("generated_images"),
+export function hasImageGenEvidence(toolCalls: ToolCall[], dirHasImage: boolean): boolean {
-  );
+  return dirHasImage || hasImageGenInvocation(toolCalls);
 }
 
 export async function verifyOutput(outputPath: string): Promise<{ bytes: number }> {

skills/baoyu-post-to-wechat/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: baoyu-post-to-wechat
 description: Posts content to WeChat Official Account (微信公众号) via API or Chrome CDP. Supports article posting (文章) with HTML, markdown, or plain text input, and image-text posting (贴图, formerly 图文) with multiple images. Markdown article workflows default to converting ordinary external links into bottom citations for WeChat-friendly output. Use when user mentions "发布公众号", "post to wechat", "微信公众号", or "贴图/图文/文章".
-version: 1.118.1
+version: 1.118.2
 metadata:
   openclaw:
     homepage: https://github.com/JimLiu/baoyu-skills#baoyu-post-to-wechat

skills/baoyu-translate/SKILL.md

@@ -6,7 +6,7 @@ description: >-
   "localize", "本地化", "refined translation", "精细翻译", "proofread translation", "快速翻译", "快翻",
   "这篇文章翻译一下", or provides a URL/file with translation intent. Supports three modes
   (quick/normal/refined) with custom glossary support.
-version: 1.59.0
+version: 1.117.3
 metadata:
   openclaw:
     homepage: https://github.com/JimLiu/baoyu-skills#baoyu-translate

skills/baoyu-wechat-summary/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: baoyu-wechat-summary
 description: Summarizes WeChat group chat highlights into a structured digest using the local wx-cli binary (https://github.com/jackwener/wx-cli). Generates a normal digest by default; a roast (毒舌) version is opt-in. Maintains per-group history (history.json + history-digests.jsonl), per-user profiles, and per-group fact memory (memory.md) across runs, with privacy guardrails baked in. Use when the user asks to "总结群聊", "群聊精华", "群聊摘要", "summarize group chat", "group chat digest", mentions a WeChat group name with a time range, says "帮我看看 XX 群最近聊了什么", "XX 群有什么值得看的", or asks to "回溯画像" / "初始化画像" / "backfill profiles". Adds the roast version when the user says "毒舌版", "roast 版", "再来个毒舌的", or similar.
-version: 1.117.3
+version: 1.117.4
 metadata:
   openclaw:
     homepage: https://github.com/JimLiu/baoyu-skills#baoyu-wechat-summary