zhayujie/chatgpt-on-wechat

Choose this fork only if you want a simpler, legacy WeChat bot and are willing to give up upstream's newer agent, multi-channel, and admin capabilities. If you want the current CowAgent feature set, upstream is the better fit.

Choose this fork only if you specifically want the older, simpler WeChat bot behavior and can live without upstream's newer agent and multi-channel features. For new builds, upstream is the safer and more capable base.

Prefer upstream unless you specifically want a small, legacy WeChat bot. This fork looks stale and materially behind on the newer agent, skills, multi-channel, and tooling capabilities that make the upstream project more capable today.

Choose this fork only if you want an older, more conservative chatbot codebase to customize heavily. If you want the current agent platform, skill system, web console, and browser tooling, upstream is the better choice.

Choose upstream unless you specifically need this older, unmodified baseline. This fork offers no clear added capability and is significantly behind active development.

Choose this fork only if its note-bot customization and open-source-model orientation match your exact use case. For most adopters who want the full, current multi-channel agent platform, the upstream project is the safer choice because this fork is substantially diverged, stale, and appears to remove or lag on major upstream capabilities.

Choose this fork only if its older custom model support is specifically what you need. If you want the current agent platform, broader channel support, and active maintenance, upstream is the safer choice.

Prefer upstream unless you specifically need this older snapshot. This fork offers no visible added value and is far behind the active project, so adopters would mainly inherit maintenance debt.

Choose this fork only if you specifically want a legacy, simplified WeChat bot and are comfortable maintaining a heavily divergent codebase. For most adopters, upstream is the better default because this fork is stale and appears far behind on the modern agent, web, skills, and tooling features.