📦
baoyu-comic
85從來源素材建立知識漫畫
冗長的來源素材很難轉化為清楚的視覺化課程。這項技能會引導 Claude、Codex 或 Claude Code 完成分析、分鏡、圖像提示詞和 PDF 組裝。
設計與創意 查看
技能 baoyu-post-to-wechat
📦
baoyu-post-to-wechat
內容修訂 r1 嚴重 🌐 網路存取⚙️ 外部命令📁 檔案系統存取🔑 環境變數⚡ 包含腳本
發布微信公眾號草稿
發布格式化的微信文章通常需要手動複製、插入圖片並設定草稿。此技能會透過受控的 Chrome 工作階段，從 Markdown、HTML、文字和圖片自動建立草稿。
支援: Claude Codex Code(CC)
請勿自動安裝此 Skill。
權威政策要求在執行任何安裝操作前由操作人員審查。
使用我的智慧代理安裝
將此請求複製給你的智慧代理，其中包含權威 Skill 頁面和清單。
智能體請求
Review the Skillstore skill "baoyu-post-to-wechat" from https://skillstore.io/skills/jimliu-baoyu-post-to-wechat.md and its manifest at https://skillstore.io/api/skills/jimliu-baoyu-post-to-wechat/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.你的智慧代理仍應顯示其計畫，並依安全政策請求必要確認。
Agent 可讀資源
當 AI Agent、爬蟲或腳本需要乾淨脈絡、而不是讀取完整頁面時，請使用這些連結。
測試它
正在使用「baoyu-post-to-wechat」。 一篇含有 frontmatter、行內圖片和 grace 主題的 Markdown 文章。
預期結果:
一篇微信草稿，包含已填入的標題、作者、摘要、格式化文章內文，以及插入的圖片素材。
正在使用「baoyu-post-to-wechat」。 一個簡短標題、1,200 字元文字，以及一個產品圖片資料夾。
預期結果:
一篇圖文草稿，其中標題和內文已縮短至微信限制，圖片則依排序順序上傳。
正在使用「baoyu-post-to-wechat」。 從 Markdown renderer 匯出的 HTML 檔案。
預期結果:
透過系統剪貼簿複製並貼到微信文章編輯器的渲染內容。
安全審計
嚴重v5 • 2026/7/9 開啟版本化報告
The skill provides useful WeChat publishing automation but relies on powerful local capabilities: Chrome DevTools control, persistent authenticated browser profiles, clipboard access, keyboard simulation, filesystem reads and writes, and external process execution. Several scanner hits are false positives from regex parsing, SVG markup, CSS, and localhost checks, but the overall behavior still requires strong user consent and sandboxing. Static review was capped at 400/677 representative findings; omitted static matches are unconfirmed, so automatic publishing stays disabled until manual review.
27
已掃描檔案
5,854
分析行數
111
複核項目
0
已忽略的誤報
已確認的安全問題 (15)
嚴重
Keylogger keywords
console.log(`Send real paste keystroke (Cmd+V / Ctrl+V) to the frontmost application
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
// Build script that activates app (if specified) and sends keystroke in one atomic operation
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
keystroke "v" using command down
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
keystroke "v" using command down
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
console.log(`[paste] Sending paste keystroke (retries=${retries}, delay=${delayMs}ms)...`);
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
console.log('[paste] Paste keystroke sent successfully');
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
console.error('[paste] Failed to send paste keystroke');
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
spawnSync('osascript', ['-e', 'tell application "System Events" to keystroke "c" using command down'
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
顯示全部 15 條已確認問題
嚴重
Keylogger keywords
spawnSync('osascript', ['-e', 'tell application "System Events" to keystroke "v" using command down'
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
Keylogger keywords
| `scripts/paste-from-clipboard.ts` | Send real paste keystroke |
The skill intentionally sends copy or paste keyboard shortcuts to the frontmost application. This is legitimate for the workflow but is high-risk UI automation because it can affect the wrong window.
嚴重
[HEURISTIC] DANGEROUS COMBINATION: Code execution + Network + Credential access
This combination is common in credential stealers and RATs
The code combines browser automation, local command execution, filesystem access, network access, and persisted browser profiles. That combination can publish content or access authenticated browser state if misused.
高
scripts/wechat-agent-browser.ts:13-17scripts/wechat-agent-browser.ts:184-195scripts/wechat-agent-browser.ts:215-217
Shell Command Injection Risk in Browser Helper
The helper builds a shell command string for agent-browser and runs it with execSync. Later calls interpolate post title, content, selector, and image data into that command path, which can allow command injection or unintended browser actions if inputs contain shell metacharacters.
execSync receives a constructed shell string, and reviewed call sites interpolate workflow-controlled strings. The exact exploitability depends on inputs, but the pattern is strong.
高
Authenticated Browser Session Automation
The skill launches Chrome with a persistent user data directory and controls an authenticated WeChat session over CDP. A compromised or misused workflow could operate inside the user account, save drafts, upload files, or interact with the editor without normal manual safeguards.
The reviewed files explicitly launch Chrome, persist a profile, connect via CDP, and wait for WeChat login. This confirms account-level browser automation intent.
中
scripts/md-to-wechat.ts:73-83scripts/md/extensions/plantuml.ts:142-145scripts/md/extensions/plantuml.ts:191-199
Remote Markdown Assets and Diagram Rendering
Markdown conversion can download remote image URLs and PlantUML rendering can send encoded diagram text to an external PlantUML server. This may disclose source content or asset URLs during article preparation.
The code directly downloads http or https images and constructs PlantUML server URLs from diagram content. This is legitimate functionality but creates clear data disclosure risk.
低
Base64 encode
const base64 = btoa(String.fromCharCode(...utf8Bytes))
Base64-style encoding is used for PlantUML diagram URLs, not malware obfuscation. It is kept as a low-confidence confirmed transformation because it sends encoded user diagram text to a server.
能力複核項目 (111)
這些是真實的本機能力，對此技能而言可能是預期行為，因此需要複核，但不會按已確認的惡意行為計分。
高
Node.js child_process module
import { spawn } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
Node.js child_process module
import { spawn } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
Node.js child_process module
import { spawnSync } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
Node.js child_process module
import { spawnSync } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
Node.js child_process module
import { execSync, spawnSync } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
Node.js child_process module
import { spawnSync } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
Node.js child_process module
import { spawn } from 'node:child_process';
The file imports child_process to launch Chrome, npx, clipboard tools, or browser helpers. External command execution is central to the workflow and increases local system risk.
高
PowerShell invocation
await runCommand('powershell.exe', ['-NoProfile', '-Sta', '-Command', ps, '-Path', imagePath]);
PowerShell is used to manipulate the Windows clipboard or send paste actions. This is expected behavior, but it can affect system state outside the skill sandbox.
高
PowerShell invocation
await runCommand('powershell.exe', ['-NoProfile', '-Sta', '-Command', ps, '-Path', htmlFilePath]);
PowerShell is used to manipulate the Windows clipboard or send paste actions. This is expected behavior, but it can affect system state outside the skill sandbox.
高
PowerShell invocation
const result = spawnSync('powershell.exe', ['-NoProfile', '-Command', ps], { stdio: 'pipe' });
PowerShell is used to manipulate the Windows clipboard or send paste actions. This is expected behavior, but it can affect system state outside the skill sandbox.
高
Process spawn
const chrome = spawn(chromePath, [
The process spawn launches local programs such as Chrome, clipboard helpers, or system tools. These calls are real external command execution and should require user awareness.
高
Process spawn
const child = spawn(command, args, { stdio: ['pipe', 'pipe', 'pipe'] });
The process spawn launches local programs such as Chrome, clipboard helpers, or system tools. These calls are real external command execution and should require user awareness.
高
Process spawn
const child = spawn(command, args, { stdio: ['pipe', 'pipe', 'pipe'] });
The process spawn launches local programs such as Chrome, clipboard helpers, or system tools. These calls are real external command execution and should require user awareness.
高
Process spawn
const chrome = spawn(chromePath, [
The process spawn launches local programs such as Chrome, clipboard helpers, or system tools. These calls are real external command execution and should require user awareness.
高
Synchronous process execution
const result = execSync(fullCmd, { encoding: 'utf8', stdio: ['pipe', 'pipe', 'pipe'] });
execSync builds and runs an agent-browser shell command. Shell-based synchronous execution is risky, especially when command text includes workflow-controlled values.
高
Synchronous spawn
const result = spawnSync('npx', ['-y', 'bun', renderScript, tempMdPath, '--theme', theme], {
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync('osascript', ['-e', script], { stdio: 'pipe' });
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync('osascript', ['-e', script], { stdio: 'pipe' });
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const which = spawnSync('which', [tool.cmd], { stdio: 'pipe' });
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync(tool.cmd, tool.args, { stdio: 'pipe' });
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync('powershell.exe', ['-NoProfile', '-Command', ps], { stdio: 'pipe' });
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync('agent-browser', ['--session', SESSION, ...args], {
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync('npx', ['-y', 'bun', copyScript, 'image', imagePath], { stdio: 'inherit' })
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
spawnSync('osascript', ['-e', 'tell application "System Events" to keystroke "c" using command down'
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
spawnSync('xdotool', ['key', 'ctrl+c']);
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
spawnSync('osascript', ['-e', 'tell application "System Events" to keystroke "v" using command down'
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
spawnSync('xdotool', ['key', 'ctrl+v']);
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
高
Synchronous spawn
const result = spawnSync('npx', args, { stdio: ['inherit', 'pipe', 'pipe'] });
The synchronous spawn runs local automation tools such as osascript, xdotool, npx, bun, or agent-browser. These commands can alter browser or clipboard state.
中
Dynamic import() expression
const { Infographic, setDefaultFont, setFontExtendFactor, exportToSVG } = await import('@antv/infogr
Dynamic import is used for optional rendering or syntax highlighting modules. One path can load language modules from a CDN, which is a supply-chain and remote-code risk.
中
Dynamic import() expression
const mermaid = await import("mermaid");
Dynamic import is used for optional rendering or syntax highlighting modules. One path can load language modules from a CDN, which is a supply-chain and remote-code risk.
中
Dynamic import() expression
const module = await import(/* @vite-ignore */ grammarUrlFor(language))
Dynamic import is used for optional rendering or syntax highlighting modules. One path can load language modules from a CDN, which is a supply-chain and remote-code risk.
中
File stream creation
fs.createReadStream(filePath).on('error', reject).pipe(child.stdin);
The code streams local files or downloaded content as part of clipboard and rendering operations. This is expected, but it handles user-selected file data.
中
File stream creation
const file = fs.createWriteStream(destPath);
The code streams local files or downloaded content as part of clipboard and rendering operations. This is expected, but it handles user-selected file data.
中
innerHTML assignment (XSS risk)
container.innerHTML = `<div style="color: red; padding: 10px; border: 1px solid red;">Infographic 渲染
The renderer writes generated HTML into the DOM. If untrusted Markdown or remote highlighting output reaches this path, it can create an XSS-style rendering risk.
中
innerHTML assignment (XSS risk)
codeBlock.innerHTML = highlighted
The renderer writes generated HTML into the DOM. If untrusted Markdown or remote highlighting output reaches this path, it can create an XSS-style rendering risk.
中
Node.js fs operations
fs.unlinkSync(destPath);
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
fs.unlinkSync(destPath);
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
fs.unlink(destPath, () => {});
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
const content = fs.readFileSync(markdownPath, 'utf-8');
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
baseCss: fs.readFileSync(basePath, "utf-8"),
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
themeCss: fs.readFileSync(themePath, "utf-8"),
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
const markdown = fs.readFileSync(inputPath, "utf-8");
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
fs.renameSync(outputPath, backupPath);
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
fs.writeFileSync(outputPath, html, "utf-8");
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
const content = fs.readFileSync(htmlPath, 'utf-8');
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
Node.js fs operations
const text = fs.readFileSync(filePath, 'utf-8');
The skill reads and writes local Markdown, HTML, image, and temporary files. This is required for publishing but exposes local file handling risk.
中
outerHTML assignment (XSS risk)
placeholderElement.outerHTML = createPlantUMLHTML(imageUrl, options, svgContent)
The PlantUML renderer can replace DOM content with SVG fetched from a remote server. Injecting remote SVG into outerHTML is a real content injection risk.
中
Synchronous file operations
const content = fs.readFileSync(markdownPath, 'utf-8');
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Synchronous file operations
baseCss: fs.readFileSync(basePath, "utf-8"),
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Synchronous file operations
themeCss: fs.readFileSync(themePath, "utf-8"),
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Synchronous file operations
const markdown = fs.readFileSync(inputPath, "utf-8");
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Synchronous file operations
fs.writeFileSync(outputPath, html, "utf-8");
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Synchronous file operations
const content = fs.readFileSync(htmlPath, 'utf-8');
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Synchronous file operations
const text = fs.readFileSync(filePath, 'utf-8');
Synchronous filesystem reads and writes process article and rendered output files. This is expected but can read or overwrite user-selected paths.
中
Temp directory access
const tempDir = await mkdtemp(path.join(os.tmpdir(), prefix));
The skill creates temporary directories and files for clipboard and article image processing. Temporary file handling is legitimate but needs cleanup and path controls.
中
Temp directory access
const tempDir = options?.tempDir ?? path.join(os.tmpdir(), 'wechat-article-images');
The skill creates temporary directories and files for clipboard and article image processing. Temporary file handling is legitimate but needs cleanup and path controls.
中
Temp directory access
"htmlPath": "/tmp/wechat-article-images/temp-article.html",
The skill creates temporary directories and files for clipboard and article image processing. Temporary file handling is legitimate but needs cleanup and path controls.
中
Temp directory access
"localPath": "/tmp/wechat-article-images/img.png",
The skill creates temporary directories and files for clipboard and article image processing. Temporary file handling is legitimate but needs cleanup and path controls.
中
WebSocket connection
const ws = new WebSocket(url);
The code connects to Chrome DevTools Protocol over WebSocket. This grants browser automation capability in an authenticated WeChat session.
中
WebSocket connection
const ws = new WebSocket(url);
The code connects to Chrome DevTools Protocol over WebSocket. This grants browser automation capability in an authenticated WeChat session.
低
Environment variable access (dot notation)
const override = process.env.WECHAT_BROWSER_CHROME_PATH?.trim();
The code reads environment variables for browser path or profile location. This appears to be configuration, but environment access is still relevant for audit visibility.
低
Environment variable access (dot notation)
const base = process.env.XDG_DATA_HOME || path.join(os.homedir(), '.local', 'share');
The code reads environment variables for browser path or profile location. This appears to be configuration, but environment access is still relevant for audit visibility.
低
Environment variable access (dot notation)
const override = process.env.WECHAT_BROWSER_CHROME_PATH?.trim();
The code reads environment variables for browser path or profile location. This appears to be configuration, but environment access is still relevant for audit visibility.
低
Environment variable access (dot notation)
const base = process.env.XDG_DATA_HOME || path.join(os.homedir(), '.local', 'share');
The code reads environment variables for browser path or profile location. This appears to be configuration, but environment access is still relevant for audit visibility.
低
Environment variable object
const override = process.env.WECHAT_BROWSER_CHROME_PATH?.trim();
The code reads process.env for browser path or data directory configuration. This is low-risk configuration access rather than credential harvesting.
低
Environment variable object
const base = process.env.XDG_DATA_HOME || path.join(os.homedir(), '.local', 'share');
The code reads process.env for browser path or data directory configuration. This is low-risk configuration access rather than credential harvesting.
低
Environment variable object
const override = process.env.WECHAT_BROWSER_CHROME_PATH?.trim();
The code reads process.env for browser path or data directory configuration. This is low-risk configuration access rather than credential harvesting.
低
Environment variable object
const base = process.env.XDG_DATA_HOME || path.join(os.homedir(), '.local', 'share');
The code reads process.env for browser path or data directory configuration. This is low-risk configuration access rather than credential harvesting.
低
Fetch API call
const res = await fetch(url, { redirect: 'follow' });
Network fetch is used for Chrome debugging endpoints, PlantUML SVG retrieval, or browser-helper logic. Network calls are real and may contact local or external services.
低
Fetch API call
const response = await fetch(svgUrl)
Network fetch is used for Chrome debugging endpoints, PlantUML SVG retrieval, or browser-helper logic. Network calls are real and may contact local or external services.
低
Fetch API call
fetch('file://${img}').then(r => r.blob()).then(b => {
Network fetch is used for Chrome debugging endpoints, PlantUML SVG retrieval, or browser-helper logic. Network calls are real and may contact local or external services.
低
Fetch API call
const res = await fetch(url, { redirect: 'follow' });
Network fetch is used for Chrome debugging endpoints, PlantUML SVG retrieval, or browser-helper logic. Network calls are real and may contact local or external services.
低
Hardcoded URL
[Link text](https://example.com)
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
const version = await fetchJson<{ webSocketDebuggerUrl?: string }>(`http://127.0.0.1:${port}/json/ve
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* https://github.com/bent10/marked-extensions/tree/main/packages/alert
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* A [marked](https://marked.js.org/) extension to support [GFM alerts](https://github.com/orgs/commu
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
// 行内公式对齐 https://groups.google.com/g/mathjax-users/c/zThKffrrCvE?pli=1
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* @default 'https://www.plantuml.com/plantuml'
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* 基于官方文档 https://plantuml.com/text-encoding
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* 基于官方文档 https://plantuml.com/text-encoding
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* 基于官方文档 https://plantuml.com/text-encoding
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
serverUrl: options.serverUrl || `https://www.plantuml.com/plantuml`,
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
* https://talk.commonmark.org/t/proper-ruby-text-rb-syntax-support-in-markdown/2279
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
Original project: https://github.com/doocs/md
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
const HLJS_CDN_BASE = `https://cdn-doocs.oss-cn-shenzhen.aliyuncs.com/npm/highlightjs/${HLJS_VERSION
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
const WECHAT_URL = 'https://mp.weixin.qq.com/';
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
const WECHAT_URL = 'https://mp.weixin.qq.com/';
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
const WECHAT_URL = 'https://mp.weixin.qq.com/';
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Hardcoded URL
const version = await fetchJson<{ webSocketDebuggerUrl?: string }>(`http://127.0.0.1:${port}/json/ve
The skill contains fixed WeChat, CDN, PlantUML, or documentation URLs. These network destinations are expected but should be transparent to users.
低
Node.js fs stat operations
if (override && fs.existsSync(override)) return override;
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (fs.existsSync(p)) return p;
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(imagePath)) throw new Error(`File not found: ${imagePath}`);
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(htmlFilePath)) throw new Error(`File not found: ${htmlFilePath}`);
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(localPath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(htmlPath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(markdownPath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(basePath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(themePath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(inputPath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (fs.existsSync(outputPath)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(img)) throw new Error(`Image not found: ${img}`);
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
} else if (htmlFile && fs.existsSync(htmlFile)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (effectiveHtmlFile && fs.existsSync(effectiveHtmlFile)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (fs.existsSync(img)) {
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (override && fs.existsSync(override)) return override;
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (fs.existsSync(p)) return p;
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(absPath)) throw new Error(`Markdown file not found: ${absPath}`);
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(absDir)) throw new Error(`Images directory not found: ${absDir}`);
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Node.js fs stat operations
if (!fs.existsSync(img)) throw new Error(`Image not found: ${img}`);
The code checks whether user-provided files, images, or browser paths exist. This is expected filesystem metadata access and should remain visible to users.
低
Temp file creation
import { mkdtemp, rm, writeFile } from 'node:fs/promises';
Temporary files are created for clipboard and rendered HTML workflows. This is expected but relevant because generated content may contain user data.
低
Temp file creation
const tempDir = await mkdtemp(path.join(os.tmpdir(), prefix));
Temporary files are created for clipboard and rendered HTML workflows. This is expected but relevant because generated content may contain user data.
風險因素
🌐 網路存取 (49)
references/article-posting.md:51 scripts/cdp.ts:69 scripts/cdp.ts:132 scripts/cdp.ts:15 scripts/cdp.ts:80 scripts/cdp.ts:18 scripts/cdp.ts:80 scripts/md/extensions/alert.ts:21 scripts/md/extensions/alert.ts:23 scripts/md/extensions/alert.ts:153 scripts/md/extensions/alert.ts:157 scripts/md/extensions/alert.ts:165 scripts/md/extensions/alert.ts:171 scripts/md/extensions/alert.ts:176 scripts/md/extensions/alert.ts:181 scripts/md/extensions/alert.ts:186 scripts/md/extensions/alert.ts:191 scripts/md/extensions/alert.ts:196 scripts/md/extensions/alert.ts:201 scripts/md/extensions/alert.ts:206 scripts/md/extensions/alert.ts:211 scripts/md/extensions/alert.ts:216 scripts/md/extensions/alert.ts:221 scripts/md/extensions/alert.ts:226 scripts/md/extensions/alert.ts:231 scripts/md/extensions/alert.ts:236 scripts/md/extensions/alert.ts:241 scripts/md/extensions/alert.ts:251 scripts/md/extensions/alert.ts:256 scripts/md/extensions/katex.ts:26 scripts/md/extensions/plantuml.ts:193 scripts/md/extensions/plantuml.ts:7 scripts/md/extensions/plantuml.ts:35 scripts/md/extensions/plantuml.ts:61 scripts/md/extensions/plantuml.ts:78 scripts/md/extensions/plantuml.ts:241 scripts/md/extensions/ruby.ts:5 scripts/md/LICENSE:3 scripts/md/utils/languages.ts:80 scripts/wechat-agent-browser.ts:195 scripts/wechat-agent-browser.ts:6 scripts/wechat-article.ts:7 scripts/wechat-browser.ts:178 scripts/wechat-browser.ts:241 scripts/wechat-browser.ts:113 scripts/wechat-browser.ts:9 scripts/wechat-browser.ts:189 scripts/wechat-browser.ts:116 scripts/wechat-browser.ts:189
⚙️ 外部命令 (50)
scripts/cdp.ts:1 scripts/cdp.ts:184 scripts/cdp.ts:70 scripts/cdp.ts:80 scripts/cdp.ts:89 scripts/cdp.ts:155 scripts/cdp.ts:182 scripts/cdp.ts:185 scripts/cdp.ts:186 scripts/cdp.ts:204 scripts/cdp.ts:223 scripts/cdp.ts:228-236 scripts/cdp.ts:240 scripts/copy-to-clipboard.ts:1 scripts/copy-to-clipboard.ts:59 scripts/copy-to-clipboard.ts:96 scripts/copy-to-clipboard.ts:11-26 scripts/copy-to-clipboard.ts:79 scripts/copy-to-clipboard.ts:108 scripts/copy-to-clipboard.ts:129-186 scripts/copy-to-clipboard.ts:215 scripts/copy-to-clipboard.ts:227 scripts/copy-to-clipboard.ts:257 scripts/copy-to-clipboard.ts:260 scripts/copy-to-clipboard.ts:273 scripts/copy-to-clipboard.ts:279 scripts/copy-to-clipboard.ts:292 scripts/copy-to-clipboard.ts:327 scripts/copy-to-clipboard.ts:372 scripts/copy-to-clipboard.ts:377 scripts/copy-to-clipboard.ts:239 scripts/copy-to-clipboard.ts:249 scripts/md-to-wechat.ts:8 scripts/md-to-wechat.ts:178 scripts/md-to-wechat.ts:44 scripts/md-to-wechat.ts:77 scripts/md-to-wechat.ts:80 scripts/md-to-wechat.ts:151 scripts/md-to-wechat.ts:164-169 scripts/md-to-wechat.ts:169 scripts/md-to-wechat.ts:177-185 scripts/md-to-wechat.ts:185-190 scripts/md-to-wechat.ts:190-213 scripts/md-to-wechat.ts:239-270 scripts/md-to-wechat.ts:270-279 scripts/md/extensions/alert.ts:116 scripts/md/extensions/alert.ts:26 scripts/md/extensions/alert.ts:36 scripts/md/extensions/alert.ts:47 scripts/md/extensions/alert.ts:48
📁 檔案系統存取 (50)
scripts/cdp.ts:35 scripts/cdp.ts:58 scripts/copy-to-clipboard.ts:260 scripts/copy-to-clipboard.ts:279 scripts/copy-to-clipboard.ts:115 scripts/copy-to-clipboard.ts:120 scripts/copy-to-clipboard.ts:3 scripts/copy-to-clipboard.ts:120 scripts/md-to-wechat.ts:35 scripts/md-to-wechat.ts:43 scripts/md-to-wechat.ts:57 scripts/md-to-wechat.ts:118 scripts/md-to-wechat.ts:79 scripts/md-to-wechat.ts:189 scripts/md-to-wechat.ts:269 scripts/md-to-wechat.ts:118 scripts/md-to-wechat.ts:28 scripts/md-to-wechat.ts:145 scripts/md-to-wechat.ts:120 scripts/md-to-wechat.ts:226 scripts/md-to-wechat.ts:230 scripts/md/extensions/alert.ts:116 scripts/md/extensions/infographic.ts:94 scripts/md/extensions/plantuml.ts:266 scripts/md/extensions/toc.ts:34 scripts/md/render.ts:557 scripts/md/render.ts:558 scripts/md/render.ts:628 scripts/md/render.ts:643 scripts/md/render.ts:646 scripts/md/render.ts:548 scripts/md/render.ts:552 scripts/md/render.ts:616 scripts/md/render.ts:641 scripts/md/render.ts:557 scripts/md/render.ts:558 scripts/md/render.ts:628 scripts/md/render.ts:646 scripts/md/render.ts:181 scripts/md/render.ts:343 scripts/wechat-agent-browser.ts:98 scripts/wechat-article.ts:149 scripts/wechat-article.ts:236 scripts/wechat-article.ts:300 scripts/wechat-article.ts:338 scripts/wechat-article.ts:149 scripts/wechat-browser.ts:18 scripts/wechat-browser.ts:133 scripts/wechat-browser.ts:167 scripts/wechat-browser.ts:298
🔑 環境變數 (8)
⚡ 包含腳本 (16)
scripts/md/extensions/alert.ts:116 scripts/md/extensions/infographic.ts:12 scripts/md/extensions/infographic.ts:77 scripts/md/extensions/infographic.ts:94 scripts/md/extensions/markup.ts:21 scripts/md/extensions/markup.ts:45 scripts/md/extensions/markup.ts:71 scripts/md/extensions/plantuml.ts:170 scripts/md/extensions/plantuml.ts:266 scripts/md/extensions/ruby.ts:31 scripts/md/extensions/ruby.ts:44 scripts/md/extensions/toc.ts:34 scripts/md/render.ts:268 scripts/md/utils/languages.ts:112 scripts/md/utils/languages.ts:202 scripts/paste-from-clipboard.ts:32
偵測到的模式
Keylogger keywords×10[HEURISTIC] DANGEROUS COMBINATION: Code execution + Network + Credential access
審計單位: claude 檢視審計紀錄 →
分享與引用此報告
分享版本化評估報告、中性徽章、嵌入卡片與引用。Skillstore 只報告證據，不判斷該 Skill 是否安全。
複製報告連結
https://skillstore.io/skills/jimliu-baoyu-post-to-wechat/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_reportMarkdown 徽章
[](https://skillstore.io/skills/jimliu-baoyu-post-to-wechat?utm_source=security_passport_badge)HTML 徽章
<a href="https://skillstore.io/skills/jimliu-baoyu-post-to-wechat?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/jimliu-baoyu-post-to-wechat/security.svg" alt="Skillstore security assessment" loading="lazy"></a>嵌入卡片
<iframe src="https://skillstore.io/embed/skills/jimliu-baoyu-post-to-wechat.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>學術引用 (APA · BibTeX · CFF)
APA 引用
JimLiu. (2026). baoyu-post-to-wechat security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/jimliu-baoyu-post-to-wechat/audits/5BibTeX 引用
@techreport{jimliu-jimliu-baoyu-post-to-wechat-2026,
author = {JimLiu},
title = {baoyu-post-to-wechat security audit report (audit version 5)},
institution = {Skillstore},
year = {2026},
number = {5},
url = {https://skillstore.io/skills/jimliu-baoyu-post-to-wechat/audits/5},
note = {Author version unspecified}
}CITATION.cff
cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "baoyu-post-to-wechat security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
- name: "JimLiu"
date-released: "2026-07-09"
url: "https://skillstore.io/skills/jimliu-baoyu-post-to-wechat/audits/5"
identifiers:
- type: other
value: "skillstore:jimliu-baoyu-post-to-wechat:audit:5"
description: "Skillstore immutable audit report identifier"
Skillstore 評分
評分依據 證據置信度: 中 64
架構
85
可維護性
85
內容
72
社群
83
規範符合度
你可以打造什麼
發布格式化文章草稿
將 Markdown 文章轉換為微信編輯器內容，包含標題、作者、摘要和行內圖片。
建立圖文更新
準備一篇簡短的微信貼文，包含壓縮後的標題、精簡內文，以及一批本機圖片。
自動化重複發布步驟
在檢閱來源檔案和圖片素材後，使用 Claude Code 或 Codex 執行發布工作流程。
試試這些提示
發布簡單的圖文草稿
使用此技能建立一篇微信圖文草稿，標題為 "{title}"，內容來自 "{summary}"，圖片來自 "{image_folder}"。不要提交。將 Markdown 轉換為文章草稿
使用 "grace" 主題將 "{markdown_file}" 轉換為微信文章草稿。插入引用的圖片，並將結果儲存為草稿。準備已檢閱的活動貼文
檢查 "{markdown_file}" 和 "{image_folder}" 是否缺少素材，然後建立微信草稿。儲存前告訴我哪些內容被縮短或變更。執行受控發布工作流程
只有在確認 Chrome 可用且我已登入後，才使用微信瀏覽器自動化。從 "{markdown_file}" 建立草稿，保持瀏覽器開啟，並且未經我批准不要點擊提交。最佳實務
- 在專用瀏覽器設定檔中執行工作流程，並在提交前檢閱草稿。
- 盡可能使用本機圖片檔案，以避免發布期間下載遠端素材。
- 首次執行時，讓微信、Chrome、剪貼簿和輔助使用權限保持可見。
避免
- 不要在未檢閱遠端圖片連結和嵌入式圖表的情況下，對不受信任的 Markdown 執行此技能。
- 需要法務、品牌或編輯核准的貼文，不要允許自動提交。
- 當敏感應用程式位於最前方時，不要使用剪貼簿自動化。
常見問題
此技能會直接發布到微信嗎？
它可以建立並儲存草稿。提交取決於命令選項，並應在核准前進行檢閱。
為什麼它需要 Chrome？
此技能會透過 Chrome 自動化和本機偵錯連線，控制微信公眾號網頁介面。
它可以在未登入的情況下運作嗎？
不行。使用者必須登入微信公眾號頁面，通常是透過掃描 QR 碼。
它會使用系統剪貼簿嗎？
是。有些文章和圖片插入步驟會將 rich HTML 或圖片複製到剪貼簿，並貼到編輯器中。
它可以處理 Markdown 中的遠端圖片嗎？
可以。轉換指令碼可以在插入前，將遠端圖片 URL 下載到暫存資料夾。
自動執行它安全嗎？
需要謹慎，因為它會啟動瀏覽器、執行外部命令、使用剪貼簿，並傳送鍵盤事件。
開發者詳細資訊
作者
JimLiu授權
MIT
Skillstore 修訂
r1
版本提示
作者未聲明版本。
引用
1ffa7643651792ccb4bd3b15d924d2c97edff755
維護新鮮度
2026/7/18
使用量
122 次下載 · 302 次瀏覽
JimLiu 的更多技能
查看全部📦
baoyu-format-markdown
83格式化 Markdown 文章
草稿通常缺乏一致的結構、frontmatter 與易讀的排版。此技能會將純文字或粗略的 Markdown 轉換成組織完善、可供審閱的 Markdown。
文件 查看
📦
baoyu-article-illustrator
80以一致的視覺風格為文章繪製插圖
長篇文章通常需要視覺素材，但選擇圖片位置和提示詞很耗時。此技能會分析 Markdown 內容、規劃插圖，並建立一致的提示詞與圖片放置位置。
設計與創意 查看
📦
baoyu-cover-image
79建立文章封面圖片
文章封面通常需要在開始生成圖片前，先有清楚的視覺概念。此技能會分析內容，並建立符合風格的封面提示詞，包含標題與版面配置指引。
設計與創意 查看
📦
release-skills
78安全發布 Baoyu Skills
手動發布可能會遺漏變更日誌、版本與標籤步驟。此技能會引導 baoyu-skills 的可重複發布工作流程，包含雙語變更日誌與本機 git 標籤。
DevOps 與基礎設施 查看
📦
baoyu-infographic
77從結構化內容建立資訊圖表
將筆記或報告轉換成清楚的視覺內容，可能需要做出許多設計決策。此技能會分析內容、推薦版面配置與風格，並為 Claude、Codex 或 Claude Code 工作流程準備圖片提示詞。
設計與創意 查看
相關技能
查看全部📦
task-state-ledger
90Track Long Tasks With a Portable State Ledger
由 glenskii
Long technical tasks lose decisions, evidence, and next actions across sessions. This skill maintains a concise Markdown state file with reviewed local evidence references.
效率 查看
📦
playwright-testing
86提升 Playwright 測試覆蓋率
由 C0ntr0lledCha0s
當設定、定位器和 fixtures 不一致時，Playwright 測試套件可能會變得不穩定。此技能提供可靠瀏覽器測試、除錯和可重用 page objects 的實用模式。
程式開發 查看
📦
webapp-testing
85建立 Playwright Web App 測試
由 AutumnsGrove
手動瀏覽器測試速度慢，且經常漏掉表單、裝置和網路狀態之間的迴歸問題。此 skill 可協助 Claude、Codex 和 Claude Code 為真實的網頁工作流程建立實用的 Playwright 測試。
程式開發 查看
📦
quiz-generator
84產生互動式課程測驗
由 92Bilal26
課程作者需要均衡的測驗來檢驗理解，而不是記憶。此 skill 會建立含 50 題的 Quiz 元件評量，並包含回饋、來源連結與答案分布檢查。
寫作與內容 查看
📦
meme-factory
84使用 Memegen 產生迷因
由 softaworks
團隊經常需要快速的視覺幽默，而不必開啟設計工具。此技能可根據提示建立迷因 URL、Markdown 嵌入內容與文字迷因格式。
設計與創意 查看
📦
nature-reader
82建立雙語論文閱讀器
由 Yuan1z0825
當翻譯、圖表、圖說與來源位置分散時，研究論文會很難研讀。這項技能會建立完整的 Markdown 閱讀成果，包含原文、中文翻譯、視覺素材，以及可追溯的錨點。
研究與學習 查看