ppx-agent (chen6896qqwee/ppx-agent) is an MCP server listed on the M8ven Trust Index. It scores 74 out of 100, grade C. It declares 33 tools. No publisher has claimed this listing.

C
Warning
74/100

ppx-agent

皮皮虾 (PPX Agent) 测试版 - 会自我修复、自我学习的超级 Agent

Warning. Serious findings were identified. Review the full report before connecting. Grades reflect the full trust pyramid: code, verification depth, and reputation. New projects cap at C until adoption is earned.

How we verified

⚡ Live Monitored: not connected

Verified is a snapshot. Live keeps it current, and builds your track record.

⚡ Connect GitHub → continuous verification on every pushwhy connect →

Who stands behind it

chen6896qqwee

Source: github_repo_search

Is this your MCP?

Claim it to get a verified publisher badge, a free copy of our full audit findings, and direct contact for any high-priority issues we find. Or connect your repo for our deepest verification, Live Monitored: read-only, revoke anytime. What we access →

Install from

The grade above is for the source repository. Registries can serve a different version, so we mark the ones we were not able to read.

// key findings
🚨
Secret credentials may flow to a network call
1 flow detected: BRAVE_API_KEY. We can’t prove the destination matches the brand the credential belongs to.
🚨
Code appears obfuscated
1 file are unreadable to a human reviewer. Cannot audit what they do.
🔐
You'll be asked for 6 credentials: FEISHU_APP_SECRET, FEISHU_VERIFY_TOKEN, PPX_AUTH_TOKEN, WECHAT_CORP_SECRET, WECHAT_ENCODING_AES_KEY, WECHAT_WEBHOOK_TOKEN
These are read from process.env at runtime. Make sure you trust where they’ll be sent.
// tools this server exposes33 tools

These names and descriptions are the publisher's own, read from the source code. We print them as written. Our assessment is the findings above, not this list.

web_search

搜索互联网, 返回网页标题+链接+摘要。

http_request

发送 HTTP 请求 (GET/POST/PUT/DELETE), 返回状态码和响应体。

notify

主动向用户通道发送通知消息 (用于长任务或异步完成提醒)。

add_schedule

添加定时任务。cron 支持 'HH:MM'(每日) 或 'after:秒数'(N秒后执行一次)。

list_schedules

列出所有定时任务。

fetch_page

抓取一个网页的正文并转成纯文本 (截断到 20000 字符)。用于读文章/文档/新闻内容后回答问题。

read_file

读取文件内容。返回文件文本。

write_file

写入文件 (覆盖)。可用于创建/修改文件。

list_dir

列出目录内容 (文件名列表)。

run_command

执行 shell 命令并返回输出。只能在工作目录内执行, 有超时。

code_act

用 Python/Node 脚本一次性完成多个操作(读文件/处理数据/写结果), 用 print/console.log 输出结果。默认关闭, 需 security.code_act=true。

get_time

获取当前日期和时间。

memory_search

搜索皮皮虾的记忆库, 返回相关事实。

memory_add

把一条重要信息写进皮皮虾的长期记忆。

read_image

读取图片文件, 返回 base64 data URL 供多模态模型理解图片内容 (需配置支持视觉的模型, 如 gpt-4o/qwen-vl/glm-4v)。

spawn_agent

派生子 agent 处理子任务并等待结果。适合需要专门角色、并行、或隔离执行的任务 (如数据分析、代码审查、多角度论证)。子 agent 共享全局经验库。支持: 单个 task; 或 tasks 数组并行派发多个子 agent + perspectives 差异化视角; arbitrate=true 时主 agent 仲裁聚合各方结果; review=true 时走 SDD 审查循环: 实施者干活 -> 只读审查者挑问题 -> 修复 -> 复审, 达上限熔断停放交主 agent 裁定 (单任务直接审查; 多任务每个子任务独立一对实施+审查, 可配 arbitrate 聚合)。

read_document

读取本地文档并转纯文本。支持 .txt/.md/.json/.csv/.html/.pdf (文字型 PDF 直接提取, 扫描件 PDF 自动 OCR)。用于读文档/报告/数据文件后回答问题。

ocr_image

识别图片或扫描件里的文字 (OCR)。需系统安装 tesseract (含中文语言包) 或配置 config.ocr 云 key。用于 read_image/read_document 读到图片却无法理解文字时。

ingest_document

读取文档, 分块后写入长期记忆 (RAG 入库), 之后可被语义检索命中。scope 用于隔离文档来源 (如 '公司制度'/'项目文档'), 避免与其他记忆混淆。

humanize

去除文本的 AI 模板腔。检查宣传腔、过度排比、模糊归因、连接词过多、句式重复、空话套话, 返回改写的自然版本。适合公众号稿、汇报、产品介绍。

write_article

分阶段写长文: 选题→结构→初稿→审稿→修改→导出。适合公众号文章、产品介绍、课程内容、需要反复修改的长文。

clarify

需求澄清: 面对模糊任务先问清需求再动手, 避免返工。传入任务描述, 返回需要澄清的问题清单; 若信息足够则直接给出执行方案。适合改代码/做项目前使用。

scene_create

创建/更新一个场景(人设)。每个场景定义智能体在这个情境下能帮用户干什么, 类似灵魂文件。可手动设定名称/介绍/能力。

scene_list

列出所有场景及其介绍/能力。

scene_describe

用 LLM 从历史对话提炼场景介绍和能力。给定场景名, 自动总结该场景的用途和能帮用户干什么。

list_capabilities

枚举当前所有可用的工具能力(category/power/enabled) 和已安装技能。用于 agent 了解自己能干啥。

enable_capability

启用一个已注册但被禁用的工具能力。name 为工具名。

disable_capability

禁用工具能力(不卸载, 可随时启用)。name 为工具名。禁用后该工具不再出现在 LLM schema 且调用被拒。

load_skill

读取一个已安装技能的 SKILL.md 全文, 供 agent 按需加载使用。id 为技能名。

create_skill

把一次成功的方法/流程沉淀为可复用的 Agent Skill。name=技能名(字母数字横线), description=一句话说明, content=SKILL.md 正文。正文推荐含三个段落(参考 addyosmani/agent-skills): ①「## 流程」逐步工作流+检查点 ②「## 反合理化」常见偷懒借口+反驳 ③「## 验证」完成后必须提供的证据。写进 skills/ 目录后自动被 loader 发现。

refine

从最近的失败工具调用轨迹自动提炼一条可复用经验教训 (自我进化闭环)。失败轨迹足够(≥2条)时, 用 LLM 提炼成一句话经验存进经验库, 后续任务自动注入上下文。

refine_skill

从最近成功的工具调用轨迹自动提炼一个可复用 Skill (自我进化闭环)。成功轨迹足够且高频工具重复出现时, 用 LLM 提炼成 skills/<name>/SKILL.md。

replay_session

从原始对话日志恢复某会话的历史(跨天/崩溃后续跑)。sessionKey=会话名(默认default), days=回溯天数(默认7), limit=返回条数(默认40)。

// environment variables
To run this server yourself, you supply these values. They go in your own MCP client configuration and stay on your machine. The secret label means the value is sensitive, not that the server mishandles it.
configFEISHU_APP_ID
🔐 secretFEISHU_APP_SECRET
🔐 secretFEISHU_VERIFY_TOKEN
configPPX_AGENT_DATA_DIR
configPPX_AGENT_GLOBAL_DATA_DIR
configPPX_AML_AUTH
configPPX_AML_AUTH_VALUE
configPPX_AML_DATA
configPPX_AML_PORT
🔐 secretPPX_AUTH_TOKEN
configPPX_DATA_DIR
configPPX_DSH_ROOTDeepSeek Harness 底座(已内嵌源码):deepseek-harness 全部架构与源码已吸收到 .deps/deepseek-harness,且 dsh 已加入 providers 首位(default_id=dsh);dsh 后端会自动定位该目录,无需再设 。未安装 dsh 依赖时健康检查会把它判为不可用并自动回退 http/cloud;安装后即作为底座生效。
configPPX_OPENCLAW_MJS可选引擎:openclaw / dsh 后端代码保留(backend: "openclaw" / "deepseek"),需自行在 config 加 provider 或用环境变量 / PPX_DSH_ROOT 指定引擎位置
configPPX_PORT
configPPX_PROVIDER
configWECHAT_AGENT_ID
configWECHAT_CORP_ID
🔐 secretWECHAT_CORP_SECRET
🔐 secretWECHAT_ENCODING_AES_KEY
🔐 secretWECHAT_WEBHOOK_TOKEN
// quality suggestions

Tool annotations

No tools have read-only/destructive annotations

Add readOnlyHint or destructiveHint annotations to every tool so hosts can warn users before invoking.

All four hints declared on every tool

33/33 tools missing one or more hints — web_search (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); http_request (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint); notify (missing: readOnlyHint, destructiveHint, idempotentHint, openWorldHint), +30 more. OpenAI's directory rejects tools where any of the four hints are missing or non-boolean.

For every tool, set all four hints (readOnlyHint, destructiveHint, idempotentHint, openWorldHint) to explicit true/false values that match the handler’s actual behaviour.

Shell command execution

1 call in production code run through a shell (src/utils/winutf8.js:10)

Prefer library functions over shell-outs. If you must shell out, ensure all inputs are properly escaped.

Readable source code

1 file are minified or bundled, which is usually build output rather than concealment

Ship unminified, readable source.

Secrets not logged

1 secret value sent to log

Redact or omit secret values from log output.

Claim the listing to review these findings one by one and send us a correction where you disagree, straight to the team. Claiming also means we tell you when the grade moves, and reach you first if we find anything urgent.

// full audit trail
The findings above are the summary. The full trail, every check we ran, each deduction, the network hosts observed and the dependency advisories, goes to verified publishers, along with an alert whenever a new one lands. Verified publishers can also review each finding and dispute it in one click. Publisher corrections have sharpened several of our checks this month, because the maintainer knows the codebase better than any scanner.
// improvement guidance — verified publishers only
We have 5 concrete improvements we can share with the publisher of this MCP. Each comes with specific guidance to raise the trust score.
// embed badge in your README
[![M8ven Verified](https://m8ven.ai/badge/mcp/chen6896qqwee/ppx-agent?variant=verified)](https://m8ven.ai/mcp/chen6896qqwee/ppx-agent)
Shows verification status without the grade. Want the grade badge instead? Remove ?variant=verified from the URL.
commit: d6991f23f5a83966be9c5fefe0bb2ae4949603eb
code hash: 7f3760ad28b732a52dccd7a795d76565af8a8b6dbaffb3357332ef0e3a8aaafc
view raw JSON →
Check MCPs from inside your assistant
Tool Check · MCP

Vetting this one by hand? Tool Check is an MCP that scores other MCPs. Add it once and ask Claude, ChatGPT, or any MCP client to grade a server, surface CVEs, check the publisher, and suggest safer alternatives — before you install.

https://m8ven.ai/api/mcp/tool-check
check_toolsearch_toolscompare_toolsrecommend_alternativescheck_publisherreport_concern
How to add it →Free · no account needed · works in any MCP client