PromptX
BLOCKgrade F · trust 53/100PromptX · 领先的AI 智能体上下文平台 | PromptX · Leading AI Agent Context Platform
Overview
From the repository's own README, as read at the audited commit.
<div align="center"> <img src="assets/logo/PromptX-transparent.png" alt="PromptX Logo" width="120" height="120"/> <h1>PromptX · Leading AI Agent Context Platform</h1> <h2>✨ Chat is all you need - Revolutionary interaction design that turns AI agents into industry experts</h2> <p><strong>Core Capabilities:</strong>AI Role Creation Platform | Intelligent Tool Development Platform | Cognitive Memory System</p> <p>Based on MCP protocol, inject professional capabilities into Claude, Cursor and other AI applications with a single command</p><!-- Badges --><p> <a href=" "><img src="https://img.shields.io/github/stars/Deepractice/PromptX?style=social" alt="Stars"/></a> <img src="https://komarev.com/ghpvc/?username=PromptX&label=views&color=0e75b6&style=flat&abbreviated=true" alt="Views"/> <a href="LICENSE"><img src="https://img.shields.io/github/license/Deepractice/PromptX?color=blue" alt="License"/></a> <a href="https://zread.ai/Deepractice/PromptX" target="_blank"><img src="https://img.shields.io/badge/Ask_Zread-_.svg?style=flat&color=00b0aa&labelColor=000000&logo=data%3Aimage%2Fsvg%2Bxml%3Bbase64%2CPHN2ZyB3aWR0aD0iMTYiIGhlaWdodD0iMTYiIHZpZXdCb3g9IjAgMCAxNiAxNiIgZmlsbD0ibm9uZSIgeG1sbnM9Imh0dHA6Ly93d3cudzMub3JnLzIwMDAvc3ZnIj4KPHBhdGggZD0iTTQuOTYxNTYgMS42MDAxSDIuMjQxNTZDMS44ODgxIDEuNjAwMSAxLjYwMTU2IDEuODg2NjQgMS42MDE1NiAyLjI0MDFWNC45NjAxQzEuNjAxNTYgNS4zMTM1NiAxLjg4ODEgNS42MDAxIDIuMjQxNTYgNS42MDAxSDQuOTYxNTZDNS4zMTUwMiA1LjYwMDEgNS42MDE1NiA1LjMxMzU2IDUuNjAxNTYgNC45NjAxVjIuMjQwMUM1LjYwMTU2IDEuODg2NjQgNS4zMTUwMiAxLjYwMDEgNC45NjE1NiAxLjYwMDFaIiBmaWxsPSIjZmZmIi8%2BCjxwYXRoIGQ9Ik00Ljk2MTU2IDEwLjM5OTlIMi4yNDE1NkMxLjg4ODEgMTAuMzk5OSAxLjYwMTU2IDEwLjY4NjQgMS42MDE1NiAxMS4wMzk5VjEzLjc1OTlDMS42MDE1NiAxNC4xMTM0IDEuODg4MSAxNC4zOTk5IDIuMjQxNTYgMTQuMzk5OUg0Ljk2MTU2QzUuMzE1MDIgMTQuMzk5OSA1LjYwMTU2IDE0LjExMzQgNS42MDE1NiAxMy43NTk5VjExLjAzOTlDNS42MDE1NiAxMC42ODY0IDUuMzE1MDIgMTAuMzk5OSA0Ljk2MTU2IDEwLjM5OTlaIiBmaWxsPSIjZmZmIi8%2BCjxwYXRoIGQ9Ik0xMy43NTg0IDEuNjAwMUgxMS4wMzg0QzEwLjY4
Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control.
claude mcp add runtime -- npx -y @agentxjs/[email protected]
{
"mcpServers": {
"runtime": {
"command": "npx",
"args": [
"-y",
"@agentxjs/[email protected]"
]
}
}
}Exposed tools (57) 52 read · 4 write · 1 destructive
Blast radius: 1 tool can delete or overwrite. An agent that can be talked into calling a tool can be talked into calling this one.
| Tool | Risk | Description |
|---|---|---|
Balanced | read | 平衡模式,系统默认行为 |
Creative | read | 创造性探索模式,广泛联想,发现远距离连接 |
DummyDriver | write | Placeholder driver for push-based event handling |
Focused | read | 聚焦检索模式,精确查找,优先常用记忆 |
concurrent-tool | read | Concurrent tool |
create_workspace_directory | write | 在工作区中创建目录(支持递归创建)。路径必须在已绑定的工作区范围内。 |
delete_workspace_item | destructive | 删除工作区中的文件或目录。目录会被递归删除。不能删除工作区根目录。路径必须在已绑定的工作区范围内。⚠️ 此操作不可逆,请确认后再调用。 |
discover | read | Discover available AI roles and tools## What It DoesLists all activatable roles and callable tools, grouped by source:- **📦 System**: Built-in PromptX roles/tools- **🏗️ Project**: Project-specific (requires \ |
example-tool | read | 示例工具 |
excel-tool | read | Excel文件读写操作工具 |
execution | read | 执行模式资源协议 |
file | read | 文件系统协议,提供本地文件访问 |
filesystem | read | 文件系统操作工具,提供读写、搜索、编辑等功能 |
http-test-tool | read | Test tool for HTTP |
input | read | Test input |
knowledge | read | 知识资源协议 |
learn | read | Load and learn PromptX resources by protocol URL## What It DoesUnified entry point for loading professional resources: role definitions, thinking models, execution skills, knowledge bases, tool manuals, and more.## When to Use- Need to use a tool but don |
list_workspace_directory | read | 列出工作区中某个目录的内容。返回文件和子目录列表(名称、绝对路径、大小、修改时间)。自动跳过隐藏文件和 node_modules 等常见忽略目录。路径必须在已绑定的工作区范围内。 |
list_workspaces | read | 获取用户绑定的工作区文件夹列表。返回每个工作区的 id、名称和绝对路径。调用此工具后,可以使用 list_workspace_directory 浏览具体目录。 |
list_xlsx_sheets | read | List all sheet names in an Excel file |
manual | read | Manual资源协议 - 加载工具使用手册和说明文档 |
pdf-reader | read | PDF分页阅读工具,支持按页码提取文本和图片 |
project | read | 项目协议,基于当前项目状态的高性能路径解析 |
prompt | read | PromptX内置提示词资源协议 |
read_docx | read | Read content from a Microsoft Word document (.docx file). Returns the text content of the document. |
read_pdf | read | Read text content from a PDF file. Returns the extracted text from all pages. |
read_pptx | read | Read content from a Microsoft PowerPoint presentation (.pptx file). Returns the text content from all slides. |
read_workspace_file | read | 读取工作区中某个文件的文本内容。支持 UTF-8 编码的文本文件。二进制文件(图片、压缩包等)不支持。大文件自动截断:最多读取前 512KB / 5000 行。路径必须在已绑定的工作区范围内。 |
read_xlsx | read | Read content from a Microsoft Excel spreadsheet (.xlsx or .xls file). Returns the data as formatted text or JSON. |
recall | read | Retrieve memories from a role |
remember | write | Save knowledge to a role |
role | read | AI角色资源协议 |
role-creator | read | 角色资源CRUD操作工具,简化角色创作流程 |
slow-tool | read | Slow tool |
stdio-test-tool | read | Test tool for stdio |
test | read | test |
thought | read | 思维模式资源协议 |
tool | read | Tool资源协议 - 加载可执行的JavaScript工具 |
tool-creator | read | 工具资源CRUD和验证工具,简化工具开发流程 |
toolx | read | 使用ToolSandbox执行通过@tool协议声明的工具 |
tracking-tool | read | Tracking tool |
user | read | 用户目录协议,直接映射到用户主目录 |
very-slow-tool | read | Very slow tool |
word-tool | read | Word文档读写修改工具 |
write_workspace_file | write | 在工作区中创建或覆盖写入文件。如果父目录不存在会自动递归创建。路径必须在已绑定的工作区范围内。⚠️ 会覆盖已有文件内容,请谨慎使用。 |
保存洞察 | read | 将重要洞察保存为记忆 |
应用记忆 | read | 检索相关经验 |
数据库报表工具 | read | 连接数据库生成报表,支持dry-run测试 |
查看使用帮助 | read | 重新学习命令使用方法 |
查看可用角色 | read | 返回角色选择页面 |
查看角色 | read | 查看可用角色列表 |
查看角色列表 | read | 选择其他角色 |
检索记忆 | read | 检索相关记忆支持思考 |
激活角色 | read | 激活完整角色能力 |
生成学习计划 | read | 为特定角色生成学习计划 |
继续学习 | read | 学习其他资源 |
继续思考 | read | 基于生成的prompt继续深化思考 |
Details
- Source
- deepractice/promptx
- npm
@agentxjs/runtime@2.0.2- Transports
- stdio · streamable-http
- License
- MIT
- Stars
- 3,688 · pushed 119d ago
Trust audit
Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- declared (10 observation(s))
- Network
- declared (6 observation(s))
- Shell
- declared (10 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (25)
const config = yaml.load(yamlInput) as any;
new Function(trimmedContent)
new Function(content)
new Function(content);
data = eval(`(${dataMatch[2]})`);icon.icns
emoji: '👨💻',
delete_workspace_item
.actrc
.eslintrc.js
.versionrc.json
.eslintrc.cjs
return crypto.createHash('md5').update(path.resolve(projectPath)).digest('hex').substr(0, 8)return crypto.createHash('md5').update(path.resolve(projectPath)).digest('hex').substr(0, 8)const hash = crypto.createHash('md5');const hash = crypto.createHash('md5');const esbuildBin = resolve(__dirname, '../../node_modules/.bin/esbuild')
translationsDir = path.join(__dirname, '../../src/main/i18n');
const devPath = path.join(__dirname, '../../../../packages/mcp-office/dist/index.js')
const nodeModulesPath = path.join(__dirname, '../../../node_modules/@promptx/mcp-office/dist/index.js')
const devPath = path.join(__dirname, '../../../../packages/mcp-workspace/dist/mcp-server.js')
"url": "http://127.0.0.1:5203/mcp"
"url": "http://127.0.0.1:5203/mcp"
"url": "http://127.0.0.1:5203/mcp"
"url": "http://127.0.0.1:5203/mcp"
Gates applied: no_behavioural_pass.
Audited 2026-09-14 · audit v0.4.0 · source sha d9f54f4d94e5 · full audit: observations/trust-audit/mcp-server/deepractice__promptx.json · Report an issue or request a re-scan
Audit history
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-09-14 | d9f54f4d94e5 | BLOCK | F | 53 | first audit |
Also audited
Servers with a similar trust profile.
Questions
What is the PromptX MCP server?
PromptX · 领先的AI 智能体上下文平台 | PromptX · Leading AI Agent Context Platform
What tools does PromptX expose?
57 in total: 52 read-only, 4 that write, and 1 that can delete or overwrite (delete_workspace_item). Every one is listed on this page with its risk.
Is PromptX safe to connect to an agent?
No — not without reading the findings first. The audit graded it F (53/100) and found 5 critical or high issues in the source. Each one is listed on this page with the file and line it is on. Separately from the audit: 1 of its tools can destroy data, so scope the token you give it to what you actually need.
What credentials does PromptX need?
No credential environment variables were found in its source, so it appears to need none.
How does PromptX run?
It speaks stdio and streamable-http, so it runs as a local process your client starts. It is published on npm as @agentxjs/runtime at 2.0.2.
How current is this page?
The grade is for one exact copy of the source (d9f54f4d94e5), read on 2026-09-14. The repository is watched and re-audited when it changes.
Provenance: OBSERVED · read 2026-09-14 · job trust-audit-2026-09-14