JevIntent
WeChat plugin that classifies a long-pressed message's intent, emotion, urgency and suggested reply posture with Jev.
Overview
JevIntent is a FkWeChat (LSPosed) plugin for Android WeChat users. Long-pressing any text message adds an 'intent' menu item; tapping it sends the message, a few lines of surrounding context and a per-chat relationship setting (friend, colleague, partner, client and so on) to the TypeSafe Jev decision model, then shows the verdict as a few local Toast notifications. Nothing is sent to the other person and chat history is not changed.
The decision model does all the reading: a 10-way intent choice, a 9-way emotion choice shown with the top probabilities, a 4-level urgency score and an 11-way reply-posture choice, plus extra questions on whether a reply is expected and how risky the message is. The README stresses that it is a judgment tool, not an auto-reply bot, because Jev produces no text. Version 1.9 starts the call as soon as the menu opens and caches results, and the README reports cold first-call latency dropping from about 1144 ms to 9 ms as a result.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Choice | What is the intent of this message? | small talk · asking for status · asking a favour · chasing progress · sales pitch · complaint · notification · venting · invitation |
| Choice | What emotion does the message carry? | calm · happy · dissatisfied · angry · anxious · in a hurry · sad · sarcastic · polite formality |
| Score | How urgent is the message? | — |
| Choice | What reply posture fits? | joking · normal · polite · praise · care · empathy · sympathy · apologise · ask for details · decline politely · do not reply yet |
How it works
- User long-presses a text message and taps the intent menu item
- Plugin sends the message, recent context and the chat's relationship preset to Jev with several typed questions
- Jev returns probabilities for intent, emotion, urgency and reply posture
- Results are shown as three fixed Toasts on the device; low top-1 confidence is a cue to read the top three
Strengths
- Runs only on the user's device UI and by default sends nothing to the chat
- Speculative pre-fetch and caching hide most API latency, per the README's measurements
- Relationship setting changes how the same sentence is interpreted
- Settings are edited as letters in a local properties file with no reload needed
Caveats
- Requires a rooted or LSPosed-capable Android setup with FkWeChat
- Message text is sent to the hosted TypeSafe API under the user's own key
- Optional modes can post results into the chat, which the other person will see
Getting started
Copy main.java and info.prop into the FkWeChat plugin folder for JevIntent, add api_key=apikey_... to config.properties, then enable and reload the plugin.At a glance
- Type
- Plugin
- Status
- beta
- Language
- Java
- Runtime
- FkWeChat / LSPosed plugin on Android
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
It is a FkWeChat plugin that analyzes a long-pressed WeChat text message with the Jev model for intent, emotion, urgency and reply posture and shows the result in local Toasts.
logicrw/awesome-jev-projects微信（FkWeChat 插件）：长按消息分析意图 / 情绪 / 回复姿态，只在本机弹提示，对方无感知.
wh000wh000/awesome-jev-live微信（FkWeChat 插件）：长按消息分析意图 / 情绪 / 回复姿态，只在本机弹提示，对方无感知.
hellogumbo/awesome-jevWeChat plugin that reads intent, tone and reply posture from a long-pressed message and shows the verdict locally. Sends nothing and changes no chat history. Chinese.
MrJev/awesome-jev
Frequently asked questions
What does JevIntent decide?
It picks the intent, emotion, urgency and a suggested reply posture for a single WeChat message, each with a probability.
Does it write replies for me?
No. Jev only returns judgments, so the plugin shows labels and probabilities, not reply text.
Does it need a TypeSafe key?
Yes. You supply your own Jev key in a local properties file; the repository ships without any key.
Related projects
DeepSeek Harness plugin that classifies each user message's emotion, urgency and department with Jev and injects it as context.
PluginbetaTypeScriptZero-dependency Android app that judges the visible WeChat message's intent, emotion, urgency and reply posture with Jev.
AppbetaJavaAndroid WeChat assistant where Jev judges the chat and scores the relationship, and a chat model drafts three replies.
AppbetaKotlinLocal chat analyser that labels each WeChat message's emotion and intent with Jev and grades how your replies landed.
ProjectprototypeJavaScript- jev-chat-windows★ 802
Windows WeChat reply assistant: local offline OCR reads the chat, Jev judges intent and tension and ranks three drafted replies to fill in.
ProjectproductionPython Android chat co-pilot that reads the active chat via Accessibility and returns typed verdicts on intent, emotion and risk; never replies.
AppbetaKotlin