context-engineering
当用户说『AI越来越笨/答非所问/忘了前面的设定』『上下文太长』『token烧太快』『RAG召回不准』，或要设计/优化 agent 的提示、记忆、检索、工具编排时使用。把『上下文窗口里到底该放什么』当成一门工程：审计相关性·冗余·预算·衰减 四维度，给出裁剪/检索/记忆/门禁四步流程与可运行脚本（上下文预算计算器+审计清单生成器）。与 prompt engineering 的区别：prompt 是写一句话，context engineering 是设计 AI 每轮看到的全部信息。触发词：上下文工程、context engineering、上下文窗口、上下文污染、上下文衰减、context ro
as observed 2026-10-03T08:46:10.219Z- Identifier
context-engineering- Source
- ClawHub
- Version observed
- 1.0.3
- Source repository
- not published
- Repository observation
- No source repository listed
- First observed here
- 2026-08-28T22:54:47.968Z
- Observations recorded
- 5
- Installs (reported upstream)
- 1
- Weekly downloads (upstream)
- 100
- Declared license
- MIT-0
Observation history
2026-10-03T08:46:10.219Z
Fields that differed: changelog latestVersion summary topics
| Field | Before | After |
|---|---|---|
changelog |
"v2.3 quality enhancement - security audit + marketing packaging + slug-content validation" | "权利层第 3 轮：① §3.2 权属宣告统一块归一——清除块内他资产事实（此前误填 uibc-core 的 DOI / commit cb6f11b / Sigstore Rekor 锚），改按本资产真值填写；② 技能件许可口径过授权清除——理论文本由 CC BY 4.0 改为「保留所有权利」，.py 依 §3.2 标 MIT；③ 保留分层许可：代码 MI |
latestVersion |
"1.0.0" | "1.0.3" |
summary |
"Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and con | "当用户说『AI越来越笨/答非所问/忘了前面的设定』『上下文太长』『token烧太快』『RAG召回不准』，或要设计/优化 agent 的提示、记忆、检索、工具编排时使用。把『上下文窗口里到底该放什么』当成一门工程：审计相关性·冗余·预算·衰减 四维度，给出裁剪/检索/记忆/门禁四步流程与可运行脚本（上下文预算计算器+审计清单生成器）。与 prompt eng |
topics |
["context","engineering"] | [] |
2026-09-21T23:20:53.794Z
Fields that differed: changelog summary topics
| Field | Before | After |
|---|---|---|
changelog |
"- Initial release of context-engineering skill.\n- Provides a structured process for auditing and optimizing the information fed into AI models, focusing on relevance, redundancy, | "v2.3 quality enhancement - security audit + marketing packaging + slug-content validation" |
summary |
"当用户说『AI越来越笨/答非所问/忘了前面的设定』『上下文太长』『token烧太快』『RAG召回不准』，或要设计/优化 agent 的提示、记忆、检索、工具编排时使用。把『上下文窗口里到底该放什么』当成一门工程：审计相关性·冗余·预算·衰减 四维度，给出裁剪/检索/记忆/门禁四步流程与可运行脚本（上下文预算计算器+审计清单生成器）。与 prompt eng | "Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and con |
topics |
[] | ["context","engineering"] |
2026-09-17T19:21:47.094Z
Fields that differed: license description
| Field | Before | After |
|---|---|---|
license |
null | "MIT-0" |
description |
"当用户说『AI越来越笨/答非所问/忘了前面的设定』『上下文太长』『token烧太快』『RAG召回不准』，或要设计/优化 agent 的提示、记忆、检索、工具编排时使用。把『上下文窗口里到底该放什么』当成一门工程：审计相关性·冗余·预算·衰减 四维度，给出裁剪/检索/记忆/门禁四步流程与可运行脚本（上下文预算计算器+审计清单生成器）。与 prompt eng | null |
2026-09-12T14:17:54.398Z
Fields that differed: changelog description summary topics
| Field | Before | After |
|---|---|---|
changelog |
"v2.3 quality enhancement - security audit + marketing packaging + slug-content validation" | "- Initial release of context-engineering skill.\n- Provides a structured process for auditing and optimizing the information fed into AI models, focusing on relevance, redundancy, |
description |
"Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and con | "当用户说『AI越来越笨/答非所问/忘了前面的设定』『上下文太长』『token烧太快』『RAG召回不准』，或要设计/优化 agent 的提示、记忆、检索、工具编排时使用。把『上下文窗口里到底该放什么』当成一门工程：审计相关性·冗余·预算·衰减 四维度，给出裁剪/检索/记忆/门禁四步流程与可运行脚本（上下文预算计算器+审计清单生成器）。与 prompt eng |
summary |
"Optimizes agent context setup. Use when starting a new session, when agent output quality degrades, when switching between tasks, or when you need to configure rules files and con | "当用户说『AI越来越笨/答非所问/忘了前面的设定』『上下文太长』『token烧太快』『RAG召回不准』，或要设计/优化 agent 的提示、记忆、检索、工具编排时使用。把『上下文窗口里到底该放什么』当成一门工程：审计相关性·冗余·预算·衰减 四维度，给出裁剪/检索/记忆/门禁四步流程与可运行脚本（上下文预算计算器+审计清单生成器）。与 prompt eng |
topics |
["context","engineering"] | [] |
Correction
If you maintain this extension and believe anything above is inaccurate, request a correction. Corrections are published, and disputed entries are marked as disputed while under review.