Repository navigation
Expand file tree
/
Copy pathposts.json
More file actions
247 lines (247 loc) · 13.1 KB
/
Copy pathposts.json
File metadata and controls
247 lines (247 loc) · 13.1 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
{
"posts": [
{
"id": "jev-到底适合做什么-一次真实业务实验后的思考",
"title": "Jev 到底适合做什么?一次真实业务实验后的思考",
"excerpt": "通过一次客服知识路由实验,分析 Jev 在准确率、成本、时延和多意图识别上的实际表现,并进一步讨论这类模型更适合放在什么样的业务决策位置。\n \n ",
"date": "2026-09-27",
"tag": "AI工程",
"file": "posts/Jev 到底适合做什么?一次真实业务实验后的思考.md",
"hash": "2ca49ddb"
},
{
"id": "vibe个工具系列03-agent-工作台-luna-mux-更新和一些设计理念",
"title": "vibe 个工具系列 03|Agent 工作台 Luna mux 更新和一些设计理念",
"excerpt": "Luna mux 从首发 v0.1.1 更新到了 v0.5.0。这次分享数据库窗格、终端交互等更新,再聊聊让人和 Agent 共用一个可视化界面的想法。\n \n ",
"date": "2026-09-12",
"tag": "AI工程",
"file": "posts/vibe个工具系列03:Agent 工作台 Luna mux 更新和一些设计理念.md",
"hash": "4d1eac85"
},
{
"id": "vibe个工具系列02-coding-agent-终端工作台luna-mux",
"title": "vibe 个工具系列 02|Coding Agent 终端工作台 Luna mux",
"excerpt": "系列第二期。终端复用器管窗口,Luna mux 想管的是项目里多个终端和多个 Agent 的协作。\n \n ",
"date": "2026-08-20",
"tag": "AI工程",
"file": "posts/vibe个工具系列02:Coding Agent 终端工作台Luna mux.md",
"hash": "34d2e2f4"
},
{
"id": "vibe个工具系列01-ssh客户端luna-remote",
"title": "vibe 个工具系列 01|SSH 客户端 Luna Remote",
"excerpt": "最近宝宝出生了,生活一下忙了不少。这次先简单分享一个自己制作的,每天都在用的软件。\n \n ",
"date": "2026-08-06",
"tag": "AI工程",
"file": "posts/vibe个工具系列01:SSH客户端Luna Remote.md",
"hash": "2cc83907"
},
{
"id": "ai-写代码会破坏架构吗-我用-nacos-做了次实测",
"title": "AI 写代码会破坏架构吗?我用 Nacos 做了次实测",
"excerpt": "我用两个 Nacos 真实需求对比三种 Codex 工作流,观察 AI 完成功能以后对应用级架构产生了什么影响。\n \n ",
"date": "2026-07-19",
"tag": "AI工程",
"file": "posts/AI 写代码会破坏架构吗?我用 Nacos 做了次实测.md",
"hash": "13b8b565"
},
{
"id": "大模型工程化发展回顾-模型越强-工程化越少吗",
"title": "大模型工程化发展回顾:模型越强,工程化越少吗?",
"excerpt": "大模型工程化随着模型能力变化不断重新分工,重心也从弥补模型能力逐渐转向管理环境、边界和反馈。\n \n ",
"date": "2026-07-05",
"tag": "AI工程",
"file": "posts/大模型工程化发展回顾:模型越强,工程化越少吗?.md",
"hash": "df3a89cf"
},
{
"id": "为什么有些大厂做不好-ai-转型",
"title": "为什么有些大厂做不好 AI 转型?",
"excerpt": "企业接入 AI 以后,很多流程还是跑不快。问题往往出在系统割裂、责任边界、隐性知识和组织协作。\n \n ",
"date": "2026-06-27",
"tag": "AI工程",
"file": "posts/为什么有些大厂做不好 AI 转型?.md",
"hash": "b0bb8892"
},
{
"id": "从代码验证到eval体系-再谈ai生成代码的质量保障",
"title": "从代码验证到 Eval 体系:再谈 AI 生成代码的质量保障",
"excerpt": "上次讨论不要逐行 Review AI 写的代码,这次继续往前走一步:把单次验证升级成可回归、可比较、可持续演进的代码质量验证体系。\n \n ",
"date": "2026-06-16",
"tag": "AI工程",
"file": "posts/从代码验证到Eval体系:再谈AI生成代码的质量保障.md",
"hash": "ad763234"
},
{
"id": "ward-agent-的一次上下文压缩策略优化",
"title": "Ward Agent 的一次上下文压缩策略优化",
"excerpt": "以 Ward 为例,记录一次把 Agent 上下文管理从最近历史和运行时摘要,推进到持久摘要、环境摘要和可观测状态的优化。\n \n ",
"date": "2026-06-11",
"tag": "Agent",
"file": "posts/Ward Agent 的一次上下文压缩策略优化.md",
"hash": "dc2407e4"
},
{
"id": "31岁-我给自己放了一个gap-month",
"title": "31岁,我给自己放了一个Gap Month",
"excerpt": "3月底,我从上一家公司离职了,记录一下在家休息这段时间的经历和思考。\n \n ",
"date": "2026-06-08",
"tag": "生活",
"file": "posts/31岁,我给自己放了一个Gap Month.md",
"hash": "8e76d078"
},
{
"id": "不要review-ai写的代码",
"title": "不要 Review AI 写的代码",
"excerpt": "大家用 AI Coding Agent 写代码之后,一个很明显的变化是,代码产出速度变快了,而且快得有点不适应。以前一个需求可能需要半天搭接口、改逻辑、补测试。现在描述清楚一点,A...\n \n ",
"date": "2026-06-02",
"tag": "AI工程",
"file": "posts/不要Review AI写的代码.md",
"hash": "d1407f19"
},
{
"id": "环境可见性-ai开发后端项目的高质量前提-以数据库为例",
"title": "环境可见性:AI开发后端项目的高质量前提——以数据库为例",
"excerpt": "后端项目里,AI 写 SQL 的主要难点在于缺少真实数据库的索引、规模和运行约束。\n \n ",
"date": "2026-05-22",
"tag": "AI工程",
"file": "posts/环境可见性-AI开发后端项目的高质量前提-以数据库为例.md",
"hash": "663452d3"
},
{
"id": "一个ai还是不够-ward-agent的team化改造",
"title": "一个 AI 还是不够:Ward Agent 的一次 Team 化改造",
"excerpt": "复杂任务需要一套能拆解、执行、验证和追踪的系统,不能只靠一个更强的模型硬扛。\n \n ",
"date": "2026-05-13",
"tag": "Agent",
"file": "posts/一个AI还是不够-Ward-Agent的Team化改造.md",
"hash": "e1164a12"
},
{
"id": "低速率大模型-api-下的agent体验优化",
"title": "低速率大模型 API 下的Agent体验优化",
"excerpt": "让 Ward 在大模型慢、速率有限、调用会失败的情况下,尽量表现得像个正常产品。\n \n ",
"date": "2026-05-12",
"tag": "AI工程",
"file": "posts/低速率大模型 API 下的Agent体验优化.md",
"hash": "984756f4"
},
{
"id": "2026年了-本地小模型能给openclaw-hermes当大脑了吗",
"title": "2026年了,本地小模型能给OpenClaw、Hermes当大脑了吗?",
"excerpt": "上一次尝试部署本地模型,还是在2025年10月,当时的体验只能说是图一乐。\n今年4月,阿里陆续开源了 Qwen3.6系列模型,其中有两款 Qwen3.6-35B-A3B 和 Qwe...\n ",
"date": "2026-04-30",
"tag": "AI技术",
"file": "posts/2026年了,本地小模型能给OpenClaw-Hermes当大脑了吗?.md",
"hash": "05219605"
},
{
"id": "ward-agent-用-harness-思路做-agent产品",
"title": "用 Harness 思路做 Agent 产品:Ward 重构实录",
"excerpt": "Ward Agent 近期做了一次较大的重构,从手写的 agent loop 迁移到了 Mini-Agent 框架。重构过程中,代码组织更清晰了,但真正让我有收获的,是用 Harn...\n \n ",
"date": "2026-04-23",
"tag": "Agent",
"file": "posts/ward-agent-用-harness-思路做-agent产品.md",
"hash": "a53a555f"
},
{
"id": "ward-更新日志-黄金监控-盘前盘后-skill接入-20260421",
"title": "Ward 更新日志:黄金监控、盘前盘后、Skill 接入",
"excerpt": "离上一篇文章过去了 10 天,Ward 陆续补上了几个之前规划的功能,做个更新记录。\n \n ",
"date": "2026-04-21",
"tag": "Agent",
"file": "posts/ward-更新日志-黄金监控、盘前盘后、skill接入-20260421.md",
"hash": "dccb569a"
},
{
"id": "ward-agent-美股市场数据分析ai工具",
"title": "被市场毒打后,我用 AI 给自己造了一个美股 Agent",
"excerpt": "我投资美股多年,回报率一直不太行。信息渠道太多,分不清哪些该看、哪些是噪音。干脆自己动手做了个工具:Ward。基于市场数据和 AI 大模型,把实时行情、技术分析、AI 解读、即时问...\n \n ",
"date": "2026-04-12",
"tag": "Agent",
"file": "posts/ward-agent-美股市场数据分析AI工具.md",
"hash": "c4a3e36b"
},
{
"id": "讲清楚-harness-一套简单可落地的-llm-agent-实践",
"title": "讲清楚 Harness:一套简单可落地的 LLM Agent 实践",
"excerpt": "LLM主要应用还是ChatBot的时候,大家在讨论 Prompt Engineering,如何设计一个精美的系统提示词,让大模型输出期望的结果。后来大家开始通过多轮会话解决复杂的问...\n \n ",
"date": "2026-04-01",
"tag": "AI工程",
"file": "posts/讲清楚 Harness:一套简单可落地的 LLM Agent 实践.md",
"hash": "0f92c650"
},
{
"id": "基于react推理模式的agent设计-以短信文案审核agent为例",
"title": "基于ReAct推理模式的Agent设计——以短信文案审核Agent为例",
"excerpt": "金融行业对外发送的短信需要严格合规审核——营销短信不能承诺保本收益、催收短信不能胁迫性表述、验证码短信必须有时效说明。\n \n ",
"date": "2026-03-22",
"tag": "Agent",
"file": "posts/基于ReAct推理模式的Agent设计--以短信文案审核Agent为例.md",
"hash": "4d4f62e2"
},
{
"id": "openclaw-workspace-架构详解-多-agent-设计指南",
"title": "OpenClaw Workspace 架构详解:多 Agent 设计指南",
"excerpt": "> 本文介绍 OpenClaw 的 Workspace 文件架构,以及如何为不同 Agent 定制个性化工作环境\n \n ",
"date": "2026-03-10",
"tag": "OpenClaw",
"file": "posts/OpenClaw Workspace 架构详解:多 Agent 设计指南.md",
"hash": "f01a70f6"
},
{
"id": "openclaw-memory-management",
"title": "OpenClaw 与 Claude 的上下文管理机制对比:一场记忆持久化的探索",
"excerpt": "> 本文基于 OpenClaw 官方文档(v2026.1.10)和 Anthropic 官方文档整理,所有技术细节均来自各自官方渠道。\n \n ",
"date": "2026-02-12",
"tag": "OpenClaw",
"file": "posts/openclaw-memory-management.md",
"hash": "5296714c"
},
{
"id": "调就是调-不调就是不调-微调是什么意思",
"title": "调就是调,不调就是不调,微调是什么意思?",
"excerpt": "这两年大模型应用的落地和发展迅速,出于本地化部署成本/性能等方面考虑,很多应用选择中小模型,结合常见的两个优化方式:微调(Fine-tuning) 和 RAG(检索增强生成),来提...\n \n ",
"date": "2025-09-30",
"tag": "AI技术",
"file": "posts/调就是调,不调就是不调,微调是什么意思?.md",
"hash": "6cb4afd8"
},
{
"id": "纸上谈兵聊context-engineering",
"title": "纸上谈兵聊 Context Engineering",
"excerpt": "> 本文前面部分由Claude根据我的零散笔记和代码生成,后面部分为本人手写总结。\n \n ",
"date": "2025-08-02",
"tag": "AI工程",
"file": "posts/纸上谈兵聊context engineering.md",
"hash": "1e4d125c"
},
{
"id": "通用mcp-host的system-prompt怎么写",
"title": "通用mcp host的system prompt怎么写",
"excerpt": "网上有一些cursor、claude code等工具的客户端架构分析和源码解读,这些工具由于大而全,会有较复杂的模块化设计、分层管理等,system prompt也包含几万字。\n这...\n ",
"date": "2025-07-26",
"tag": "MCP",
"file": "posts/通用mcp host的system prompt怎么写.md",
"hash": "61862cae"
},
{
"id": "mcp协议完整落地案例",
"title": "MCP协议完整落地案例",
"excerpt": "> 本文由Kimi k2和本人合作编写。\n \n ",
"date": "2025-07-20",
"tag": "MCP",
"file": "posts/MCP协议完整落地案例.md",
"hash": "cb5ad7e5"
},
{
"id": "困惑到实践-mcp的几个细节探究",
"title": "困惑到实践:MCP的几个细节探究",
"excerpt": "1.有些MCP工具可以操作本地文件和UI,有些又可以提供公共的互联网产品服务(例如高德地图),这些都是MCP Server的能力吗?MCP Server之间有什么区别?\n2.MCP...\n ",
"date": "2025-07-12",
"tag": "MCP",
"file": "posts/困惑到实践:MCP的几个细节探究.md",
"hash": "e3a5174f"
}
]
}