AGPL-3.0 local-first WeChat analysis platform whose bundled DSH plugin lets DeepSeek Harness query chat data by date or keyword once you grant access.
DSH integration
Compatible
Author-claimed
Safety audit
Unaudited
Last verified
2026-11-12
License
AGPL-3.0
01What can it help you accomplish?
Ask WeLink chat data a question from inside DeepSeek Harness instead of opening WeLink in a browser
Date-based or keyword answers covering both private and group chats, with no chat text exposed until you grant a session
DeepSeek Harness users who already run a WeLink backend and want chat lookups in the agent session
Analyse a WeChat chat history locally with AI, from statistics through relationship reports
A local Go + React app that indexes your decrypted WeChat databases and answers with dashboards, AI analyses, memories and exports
Users who have exported their own WeChat databases and want local-first analysis
02How to install into DeepSeek Harness
Prerequisites
- A decrypted WeChat database directory (`decrypted/` with `contact/contact.db` and `message/message_N.db`) — WeLink does not decrypt it for you
- For the DSH plugin: a WeLink backend build that includes the Agent query API, plus the plugin source under `plugins/dsh-welink`
- Docker Compose for the container path, or macOS 12+ / Windows 10 1903+ for the desktop app
Installation steps
- 01
Decrypt the WeChat databases with a tool compatible with your OS and WeChat version, producing a `decrypted/` directory
- 02
Docker path: put `decrypted/` in the repository root, run `docker compose up`, then open `localhost:3418`
- 03
Desktop path: download the app from GitHub Releases and pick the `decrypted/` directory in its settings page
- 04
DSH path: use the WeLink DSH plugin from `plugins/dsh-welink` against a backend that exposes the Agent query API, then grant the sessions you want the agent to see
Verify the integration
Not specified by the author
03DSH integration and capability boundaries
WeLink DSH plugin shipped with the repository source; once configured it lets DeepSeek Harness query WeLink chat data by date or keyword, against a WeLink backend that exposes the Agent query API
DSH chat query with explicit authorisation
A date or a keyword, in a session you have authorised→Private and group chat results surfaced to the DSH agent; nothing is exposed until authorisation is configured
The plugin authorises no session and opens no chat text by defaultAgent-side summarising of chat lookups
Messages the agent retrieved for a date→Structured output such as to-dos, owners and items still to confirm, produced inside the DSH session
The demo in the README uses fictional group-chat dataLocal-first analysis app
Decrypted WeChat databases→A single Go binary with embedded frontend (Docker image under 50 MB) or a macOS / Windows desktop app
The databases are decrypted, indexed and searched on your own machine; snippets only leave it when you explicitly trigger an AI conversation
04Who is it for? When not to use it?
Good for
- DeepSeek Harness users who already run a WeLink backend and want chat lookups in the agent session
- Users who have exported their own WeChat databases and want local-first analysis
Not for
- WeLink is local-first, but cloud AI, a remote training machine or a cloud export sends the relevant content to the service you choose, so check the data scope and privacy settings first. WeLink does not decrypt the WeChat database itself.
- The DSH plugin ships with the source and needs a WeLink backend that includes the Agent query API; it is not a standalone npm install.
- WeLink is AGPL-3.0, which matters if you plan to run a modified version as a network service.
- The original `ylytdeng/wechat-decrypt` repository is unavailable; the README points at a community fork but says compatibility of its output with WeLink is unverified.
05Compatibility, maintenance and safety notes
- WeLink is local-first, but cloud AI, a remote training machine or a cloud export sends the relevant content to the service you choose, so check the data scope and privacy settings first. WeLink does not decrypt the WeChat database itself.
- The DSH plugin ships with the source and needs a WeLink backend that includes the Agent query API; it is not a standalone npm install.
- WeLink is AGPL-3.0, which matters if you plan to run a modified version as a network service.
- The original `ylytdeng/wechat-decrypt` repository is unavailable; the README points at a community fork but says compatibility of its output with WeLink is unverified.
AGPL-3.0 source; every claim here comes from the repository README at the recorded commit, with no independent safety audit.
06Frequently asked questions
How do I get the WeLink plugin into DeepSeek Harness?
The plugin ships with the WeLink source under `plugins/dsh-welink`. It needs a WeLink backend that includes the Agent query API, so it is not a standalone package install — run the backend, then install the plugin from the repository source.
Does the plugin read my chats automatically?
No. The README says the plugin authorises no session and does not open chat text by default; only after you configure authorisation can DSH query by date or keyword, across private and group chats.
What do I need before WeLink can run at all?
A decrypted WeChat database directory containing `contact/contact.db` and `message/message_N.db`. WeLink does not decrypt it — you need a tool compatible with your OS and WeChat version, and the README warns that the community fork it points at has not been verified for output compatibility.
Where does my data go?
Decryption, indexing and retrieval happen on your own machine. Content only leaves it when you explicitly trigger a cloud AI conversation, a remote training machine or a cloud export — then the relevant content goes to the service you chose.
What licence does WeLink use?
AGPL-3.0. That matters if you intend to expose a modified version as a network service.
07Related DSH workflows
weknora
by tencent
Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
honcho
by plastic-labs
Memory library for building stateful agents
webnovel-writer
by lingfengqaq
A long-form Chinese web-novel writing assistant for Claude Code and DeepSeek Harness, designed to reduce memory loss and hallucinations in serialized projects of up to 2 million characters.
reme
by agentscope-ai
ReMe: Memory Management Kit for Agents - Remember Me, Refine Me.
08Data and sources
使用 DeepSeek Harness 的用户也可以安装 [WeLink DSH 插件](plugins/dsh-welink/README.md)。插件默认不授权任何会话、不开放聊天原文；配置授权后可在 DSH 中按日期或关键词查询，支持…
*演示使用虚构群聊数据：DSH 查询指定日期的消息，并整理出待办、负责人和待确认事项。*
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-11-12. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.