DATA CUT 2026-07-27 116 活跃模型 680 A 类案例

CASE EVIDENCE / A RECORD

MisterKarott 使用 GLM-5 处理软件工程任务执行

MisterKarott 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T09:14:54Z。

原始记录:glm-quota tracks GLM-5 family usage and context inside Claude Code statuslines

A

Chinese Brief

中文案例导读

MisterKarott 公开的代码代理与软件工程案例,来源为 公开代码库,复核于 2026-06-27T09:14:54Z。 Model Atlas 将它标记为 A 类证据,因为它同时具备具体使用者、具体任务、公开原始证据和可访问产物。 Model Atlas 不把 benchmark、教程、发布说明或集合页包装成真实案例。

厂商Z AI / GLM
模型GLM-5
任务类型代码代理与软件工程
审核状态auto_approved

任务

真实任务背景

这是一个围绕软件工程任务执行的真实任务,公开材料可以回溯到具体使用者和具体产物。 原始资料写作:glm-quota is a Claude Code plugin that activates when developers run Claude Code with Z.ai/GLM, displaying the current model, context-window usage, token counts, compacting reminders, Z.ai quota windows, and MCP tool-ca…

代码代理与软件工程公开代码库A 类可核验real_case
公开产物

公开材料提供公开代码、README 或项目配置,可用于核验任务结果、项目形态和模型绑定关系。 原始资料写作:The public repo documents installation through a Claude plugin marketplace or manual clone and shows a statusline example for a GLM-5.1 run with 1M context, context percentage, token counts, quota reset timers, and MCP …

模型作用

GLM-5 在该案例中承担软件工程任务执行相关的生成、分析、编排或实现角色。 原始资料写作:GLM-5 family models are the coding backend whose context and quota usage the plugin observes; the plugin turns model-runtime limits into actionable developer feedback during coding sessions.

风险边界

当前判断基于公开材料;若产物下线、仓库变更或模型参与比例仅来自作者自述,需要在引用前重新复核。 原始资料写作:Evidence specifically shows GLM-5.1 in the statusline example; treated as GLM-5-family evidence. The plugin observes usage rather than generating end-user content itself.