context-mode 是一个「skill」类 AI agent,解决:通过沙盒输出、会话记忆持久化和跨平台路由，降低AI编码代理的上下文开销。定价:产品页未标明。截至 2026-10-11,KanonAgent 记录到 4.3k 票。本站首次收录于 2026-07-14。
优化 AI 编码代理上下文窗口的工具，通过沙盒输出、会话记忆持久化和跨平台路由，显著降低上下文开销。
4.3k 票
Kanon 于 2026 年 7 月 14 日 收录
状态依据:npm_downloads_7d momentum fell to -4420 per 3d from 369 (<= 40% of the prior window)
📈 时间线什么时候变了什么
- 降温 npm_downloads_7d momentum fell to -4420 per 3d from 369 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to -5945 per 3d from 971 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to 0 per 3d from 472 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 369 in the last 3d vs 103 in the 3d before (ratio 3.6x, threshold 1.5x)
- 降温 npm_downloads_7d momentum fell to 0 per 3d from 472 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to 472 per 3d from 4122 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to 560 per 3d from 3665 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to -42 per 3d from 4279 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 4122 in the last 3d vs 1341 in the 3d before (ratio 3.1x, threshold 1.5x)
- 降温 npm_downloads_7d momentum fell to 3665 per 3d from 11437 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 4122 in the last 3d vs 1341 in the 3d before (ratio 3.1x, threshold 1.5x)
- 降温 npm_downloads_7d momentum fell to 2807 per 3d from 11668 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to 1341 per 3d from 10941 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 11437 in the last 3d vs 845 in the 3d before (ratio 13.5x, threshold 1.5x)
- 降温 npm_downloads_7d momentum fell to 1341 per 3d from 10941 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 11668 in the last 3d after a flat 3d baseline (no prior movement to compare a ratio against)
- 异常 daily change 10096.0 is 7.5 standard deviations from this entity's own prior mean (-160.8±1367.3, baseline excludes this point) — unusual, not necessarily adoption
- 异常 daily change 4985.0 is 3.8 standard deviations from this entity's own prior mean (-160.8±1367.3, baseline excludes this point) — unusual, not necessarily adoption
- 降温 npm_downloads_7d momentum fell to -5686 per 3d from 4248 (<= 40% of the prior window)
- 降温 npm_downloads_7d momentum fell to -5594 per 3d from 4248 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 4248 in the last 3d after a flat 3d baseline (no prior movement to compare a ratio against)
- 降温 npm_downloads_7d momentum fell to 0 per 3d from 988 (<= 40% of the prior window)
- 加速 npm_downloads_7d gained 4248 in the last 3d vs 988 in the 3d before (ratio 4.3x, threshold 1.5x)
- 降温 npm_downloads_7d momentum fell to 0 per 3d from 5182 (<= 40% of the prior window)
🔎 已知 / 未知
23 个字段里 15 个未知
| 可机器调用 | — | 未知 | — |
| 开源 | 1 | 已核实 | 1 条引文 |
| 可自托管 | — | 未知 | — |
| 需自带 key | — | 未知 | — |
| 自主度 | 2 | 推断 | 2 条引文 |
| 计费模式 | — | 未知 | — |
| 集成面 | mcp | 推断 | 1 条引文 |
「未知」= 我们没核实过,**不等于「否」**。硬性筛选永远不会把未知当成否。
🤖 Agent 拆解 · skill
解决什么场景通过沙盒输出、会话记忆持久化和跨平台路由，降低AI编码代理的上下文开销
自主度L2 · 有工具调用(引文: "Sandboxes tool output (98% reduction)…")
目标市场AI编码用户
前置条件开源
集成面mcp
牵引 · 为什么在涨40
为什么值得关注
解决了 AI 编码中上下文溢出与状态丢失的核心痛点，是提升 AI 代理长期任务表现的关键基础设施。
证据引文判定所依据的原文 · 逐字引用
“public GitHub repository (README fetched)”— structural
“Sandboxes tool output (98% reduction)”— description
“persists session memory”— description
“enforces routing across 17 platforms via MCP + hooks.”— description
“claude-code-hooks, claude-code-plugins, claude-code-skill, mcp-server, skills”— topics
信号来源: GitHub
访问官网 →
📛 官方徽章挂到官网 / README
你是 context-mode 的作者?上面选样式,嵌入代码实时更新;要贴合站点配色可在 URL 上加 bg= / fg= / accent=(十六进制色)。徽章链回本页。
手机端点「分享」直达微信/朋友圈/小红书;桌面端用「复制文案」后到 App 内粘贴发布
常见问题
context-mode 是什么?
优化 AI 编码代理上下文窗口的工具，通过沙盒输出、会话记忆持久化和跨平台路由，显著降低上下文开销。
context-mode 解决什么问题?
通过沙盒输出、会话记忆持久化和跨平台路由，降低AI编码代理的上下文开销
context-mode 为什么值得关注?
解决了 AI 编码中上下文溢出与状态丢失的核心痛点，是提升 AI 代理长期任务表现的关键基础设施。
context-mode 是开源的吗?
是。context-mode 为开源项目。
context-mode 能接哪些工具?
mcp
context-mode 有多少人在用?
KanonAgent 记录到:4.3k 票(本站首次收录于 2026-07-14)。
context-mode 有什么替代品?
KanonAgent 库内的同类 agent:Stealth Venture、orca、open-design、deer-flow、ruflo、career-ops。