phamcucpqrwbbnposhmn96-ui/KSNStudy
Productivity & UtilitiesAn injected module for the Chaoxing Learning app that automates answering, check-ins and background keep-alive on Android.
What it is
KSNStudy is a module that injects itself into the process of Chaoxing's 「学习通」 Android app. It attaches to the app's WebView and drives the chain of question recognition, answer generation, page interaction and check-in. It does not need root or an emulator, and reuses the official app's account and session.
The project combines two layers: an injected Java/DEX engine loaded through a child-first DexClassLoader, and a bundled local inference engine for answer generation. An optional OpenAI-compatible endpoint can be configured instead. Besides auto-answering, it covers automatic submission, foreground-service keep-alive during screen-off, auto check-in (location, code, face, gesture), GitHub self-update, and optional video progress and course auditing borrowed from xxt-sign-android.
The README documents the injection bridge, the bitmask probe used to connect the JS bridge reliably, wake-lock and WebView keep-alive logic, and several exam anti-cheat detection points that are neutralized in the app's smali layer.
Getting started
The distribution is an APK of roughly 240 MB published in Releases. Steps as described:
- Download the latest Release APK and wait for the full download.
- Allow installation from unknown apps (on VIVO: Settings → Privacy/Permissions).
- Install over the existing Chaoxing app — it uses the same signature, so data is retained.
- Open Chaoxing; an 「学习通 · 学习助手」 panel appears under About.
- Fill in your own OpenAI-compatible
endpoint,apiKeyandmodelin the assistant settings (see AI configuration).
Example settings shown in the README: endpoint = https://api.example.com/v1, apiKey = sk-your-key-here, model = gpt-4o-mini. These are placeholders; you must supply credentials you are authorized to use. Answers can then be generated on-device or through your endpoint.
Constraints
- Android 7.0+ on arm64 or 32-bit; some devices (e.g. VIVO) lack
addWebMessageListener, in which case the dual bridge degrades to JSI-only. - No API key is bundled; you need your own OpenAI-compatible service. Settings live in the app's private directory and survive upgrades but not a fresh install.
- Engine DEX files, native libraries and build artifacts are not in the repository — only source and docs are published, so the APK cannot be reproduced from the repo alone as described.
- Installing over Chaoxing requires matching the app's signature; the README states this is a same-signature upgrade.
- The module modifies Chaoxing's smali-layer anti-cheat checks, uses a locked coordinate with
mockDatafor remote check-in, and auto-submits answers. The README frames this as research use and puts the consequences on the user; it may conflict with the platform's terms of service. - The repository declares MIT, while GitHub reports NOASSERTION for the license metadata.
HotRepo's take
The repository appeared on the Java new-project list on 2026-11-15 with 12 stars gained over a one-hour window, reaching 114 stars and 7 forks. That is a very short observation window, so the number says little about sustained interest. What stands out technically is the unusually specific write-up of injection and self-update failure modes — the bitmask bridge probe, the FileProvider Uri.toString() pitfall, and the truncated-APK EOCD check — which makes the README more useful as a record of Android injection problems than as installation guidance. The main artifact is still a large prebuilt APK that rewrites another vendor's app and suppresses its anti-cheat checks, so readers should weigh the distribution model before trying it.
Written by AI from the project's README, updated . The project's own docs are the source of truth.