Skip to content

Commit 603b864

Browse files
committed
Enhance documentation with new entries and updates
- Added new tools to the AI Awesome section, including Unstructured-IO and MODSetter. - Updated AI Dev Awesome with new resources for open-source projects and task runners. - Expanded ComfyUI documentation with detailed model directory structure and usage benefits. - Introduced new entries in the agent documentation and MCP notes. - Added new psychology concepts including MBTI and OCEAN models. - Enhanced Grafana API documentation with examples and data frame design insights. - Updated various service and observability notes with additional resources and clarifications.
1 parent 330e326 commit 603b864

File tree

29 files changed

+524
-46
lines changed

29 files changed

+524
-46
lines changed

notes/ai/ai-awesome.md

Lines changed: 7 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -36,6 +36,13 @@ tags:
3636
- MIT, TS
3737
- MiniPerplx -> scira
3838
- minimalistic AI-powered search engine
39+
- Tools
40+
- [Unstructured-IO/unstructured-api](https://github.com/Unstructured-IO/unstructured-api)
41+
- Apache-2.0, Python
42+
- [Unstructured-IO/unstructured](https://github.com/Unstructured-IO/unstructured)
43+
- [MODSetter/SurfSense](https://github.com/MODSetter/SurfSense)
44+
- Apache-2.0, Python, TS
45+
- Open Source Alternative to NotebookLM / Perplexity, connected to external sources such as Search Engines, Slack, Linear, Jira, ClickUp, Confluence, Notion, YouTube, GitHub
3946
- Workflow/Research
4047
- [deepset-ai/haystack](https://github.com/deepset-ai/haystack)
4148
- [langgenius/dify](https://github.com/langgenius/dify)

notes/ai/dev/agent/agent-awesome.md

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -8,3 +8,6 @@ tags:
88
- https://github.com/a2aproject/A2A
99
- https://github.com/microsoft/autogen
1010
- https://github.com/DavidZWZ/Awesome-Deep-Research
11+
- https://github.com/siteboon/claudecodeui
12+
- https://github.com/eyaltoledano/claude-task-master
13+
- https://github.com/sugyan/claude-code-webui

notes/ai/dev/ai-dev-awesome.md

Lines changed: 8 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -113,7 +113,8 @@ tags:
113113
- Build AI-powered applications with React, Svelte, Vue, and Solid
114114
- https://ai-sdk.dev/providers/openai-compatible-providers
115115
- [moeru-ai/xsai](https://github.com/moeru-ai/xsai)
116-
- https://github.com/openai/openai-agents-js
116+
- [openai/openai-agents-js](https://github.com/openai/openai-agents-js)
117+
- MIT, TS
117118
- Research/Planning/Agentic/Agent/智能体
118119
- [mannaandpoem/OpenManus](https://github.com/mannaandpoem/OpenManus)
119120
- MIT, Python
@@ -189,6 +190,12 @@ tags:
189190
- 识别 UI 交互元素
190191
- Pure Vision Based GUI Agent
191192
- Tasks/特定任务
193+
- [Fission-AI/OpenSpec](https://github.com/Fission-AI/OpenSpec)
194+
- 比 spec-kit 好点,
195+
- 适合已有项目使用
196+
- [github/spec-kit](https://github.com/github/spec-kit)
197+
- 适合从 0-1
198+
- 太过于繁杂
192199
- [andrewyng/translation-agent](https://github.com/andrewyng/translation-agent)
193200
- Agentic translation using reflection workflow
194201
- [kingjulio8238/memary](https://github.com/kingjulio8238/memary)

notes/ai/dev/comfyui.md

Lines changed: 53 additions & 23 deletions
Original file line numberDiff line numberDiff line change
@@ -94,28 +94,28 @@ No module named pip
9494
9595
### ComfyUI Models 目录结构详解
9696
97-
| dir | for |
98-
| :----------------- | :-------------------------------------------------------------------------------------------------------------------------------------------------------- |
99-
| `checkpoints` | 核心基础模型,也叫“大模型”。这是文生图的起点,决定了图像生成的基础风格和能力。例如 Stable Diffusion v1.5, SDXL, 以及社区训练的各种整合模型。 |
100-
| `loras` | LoRA 模型。这些是小型微调文件,用于向基础模型添加特定的角色、画风、概念或服装,灵活性极高。 |
101-
| `vae` | VAE 模型。用于图像的编码和解码。独立的 VAE 文件可以修正图像的色彩(如改善灰蒙蒙的问题)或修复手部等细节问题。SDXL 模型通常不需要额外 VAE。 |
102-
| `controlnet` | ControlNet 模型。用于精确控制图像的生成,例如通过姿势骨架、深度图、线稿、二维码等来引导构图和内容。 |
103-
| `upscale_models` | 图像放大模型。用于“图像放大 (模型)”节点,提升图片分辨率并优化细节。例如 ESRGAN, SwinIR, 4x-UltraSharp 等。 |
104-
| `embeddings` | 文本反演 (Textual Inversion) 嵌入,也叫 Embedding。这些是极小的文件,通过一个关键词触发特定的概念、角色或画风。常用于负面提示词(如 `bad-hands-5`)。 |
105-
| `clip` | CLIP 文本编码器模型。通常 ComfyUI 会自动从大模型中加载,但你也可以把独立的 CLIP 模型放在这里,供高级工作流使用。 |
106-
| `clip_vision` | CLIP Vision 模型。用于分析图像内容,是 IPAdapter、PhotoMaker 等“图像提示”功能的核心组件。 |
107-
| `style_models` | 风格模型。主要用于 T2I-Adapter,功能与 ControlNet 类似,但更侧重于风格的迁移。 |
108-
| `hypernetworks` | Hypernetwork 模型。一种比 LoRA 更早出现的微调技术,现在已不常用,但 ComfyUI 仍然支持加载。 |
109-
| `unet` | U-Net 模型。U-Net 是 Stable Diffusion 模型的核心降噪网络。普通用户几乎不会用到这个目录,主要用于模型开发和研究,将 U-Net 单独分离出来加载。 |
110-
| `text_encoders` | 文本编码器模型。与 `unet` 类似,用于模型研究,允许单独加载和替换文本编码器部分。 |
111-
| `photomaker` | PhotoMaker 模型。一种专门用于根据输入人脸照片生成统一角色的模型。 |
112-
| `sams` | SAM (Segment Anything Model) 模型。由 Meta 开发的图像分割模型,在 ComfyUI 中用于精确地创建和分离遮罩 (Mask)。 |
113-
| `gligen` | GLIGEN 模型。用于“限定区域生成”,允许你通过画框来指定某个物体在图像中的特定位置和大小。 |
114-
| `diffusers` | 用于存放 Hugging Face 的 Diffusers 格式模型。这种格式是一个包含多个子目录和文件的文件夹,而不是单个文件。ComfyUI 可以直接加载这种格式。 |
115-
| `configs` | 配置文件。存放一些旧的 `.ckpt` 模型所需要的 `.yaml` 配置文件,以帮助 ComfyUI 识别其模型架构(如 v1 或 v2)。现在的 `.safetensors` 模型通常不需要。 |
116-
| `vae_approx` | VAE 近似解码器模型。这些是极小的、速度极快的模型,用于在 KSampler 采样过程中生成快速预览图,而不是每次都调用完整的 VAE。 |
117-
| `onnx` | ONNX 模型。用于存放已转换为 ONNX (Open Neural Network Exchange) 格式的模型,通常用于在非 NVIDIA 硬件(如 AMD 显卡)上通过 DirectML 或 Olive 进行推理。 |
118-
| `diffusion_models` | 扩散模型组件。一个更通用的目录,类似于 `unet`,用于存放扩散模型的某些部分。主要供模型开发者使用。 |
97+
| dir | for |
98+
| :----------------- | :----------------------------------------------------------------------------------------------------------------------------------------------------- |
99+
| `checkpoints` | 核心基础模型,也叫“大模型”。这是文生图的起点,决定了图像生成的基础风格和能力。例如 Stable Diffusion v1.5, SDXL, 以及社区训练的各种整合模型。 |
100+
| `loras` | LoRA 模型。这些是小型微调文件,用于向基础模型添加特定的角色、画风、概念或服装,灵活性极高。 |
101+
| `vae` | VAE 模型。用于图像的编码和解码。独立的 VAE 文件可以修正图像的色彩(如改善灰蒙蒙的问题)或修复手部等细节问题。SDXL 模型通常不需要额外 VAE。 |
102+
| `controlnet` | ControlNet 模型。用于精确控制图像的生成,例如通过姿势骨架、深度图、线稿、二维码等来引导构图和内容。 |
103+
| `upscale_models` | 图像放大模型。用于“图像放大 (模型)”节点,提升图片分辨率并优化细节。例如 ESRGAN, SwinIR, 4x-UltraSharp 等。 |
104+
| `embeddings` | 文本反演 (Textual Inversion) 嵌入,也叫 Embedding。这些是极小的文件,通过一个关键词触发特定的概念、角色或画风。常用于负面提示词(如 `bad-hands-5`)。 |
105+
| `clip` | CLIP 文本编码器模型。通常 ComfyUI 会自动从大模型中加载,但你也可以把独立的 CLIP 模型放在这里,供高级工作流使用。 |
106+
| `clip_vision` | CLIP Vision 模型。用于分析图像内容,是 IPAdapter、PhotoMaker 等“图像提示”功能的核心组件。 |
107+
| `style_models` | 风格模型。主要用于 T2I-Adapter,功能与 ControlNet 类似,但更侧重于风格的迁移。 |
108+
| `hypernetworks` | Hypernetwork 模型。一种比 LoRA 更早出现的微调技术,现在已不常用,但 ComfyUI 仍然支持加载。 |
109+
| `unet` | U-Net 模型。U-Net 是 Stable Diffusion 模型的核心降噪网络。普通用户几乎不会用到这个目录,主要用于模型开发和研究,将 U-Net 单独分离出来加载。 |
110+
| `text_encoders` | 文本编码器模型。与 `unet` 类似,用于模型研究,允许单独加载和替换文本编码器部分。 |
111+
| `photomaker` | PhotoMaker 模型。一种专门用于根据输入人脸照片生成统一角色的模型。 |
112+
| `sams` | SAM (Segment Anything Model) 模型。由 Meta 开发的图像分割模型,在 ComfyUI 中用于精确地创建和分离遮罩 (Mask)。 |
113+
| `gligen` | GLIGEN 模型。用于“限定区域生成”,允许你通过画框来指定某个物体在图像中的特定位置和大小。 |
114+
| `diffusers` | 用于存放 Hugging Face 的 Diffusers 格式模型。这种格式是一个包含多个子目录和文件的文件夹,而不是单个文件。ComfyUI 可以直接加载这种格式。 |
115+
| `configs` | 配置文件。存放一些旧的 `.ckpt` 模型所需要的 `.yaml` 配置文件,以帮助 ComfyUI 识别其模型架构(如 v1 或 v2)。现在的 `.safetensors` 模型通常不需要。 |
116+
| `vae_approx` | VAE 近似解码器模型。这些是极小的、速度极快的模型,用于在 KSampler 采样过程中生成快速预览图,而不是每次都调用完整的 VAE。 |
117+
| `onnx` | ONNX 模型。用于存放已转换为 ONNX (Open Neural Network Exchange) 格式的模型,通常用于在非 NVIDIA 硬件(如 AMD 显卡)上通过 DirectML 或 Olive 进行推理。 |
118+
| `diffusion_models` | 扩散模型组件。一个更通用的目录,类似于 `unet`,用于存放扩散模型的某些部分。主要供模型开发者使用。 |
119119
120120
**AI Art**
121121
@@ -212,7 +212,7 @@ No module named pip
212212
## API
213213
214214
```bash
215-
215+
[object Object]
216216
```
217217
218218
## 参考 {#reference}
@@ -299,6 +299,36 @@ For debugging consider passing CUDA_LAUNCH_BLOCKING=1
299299
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.
300300
```
301301
302+
## Why ComfyUI
303+
304+
**优点**
305+
306+
- **速度与灵活性**
307+
ComfyUI 可以很快地搭建和调整流程,没有固定套路,想怎么用都行。
308+
- **过程可视化**
309+
整个生成流程都能看到,每个节点做什么一目了然,方便理解图片是怎么一步步生成的。
310+
- **分享和协作简单**
311+
搭好的流程可以很容易分享给别人,也能导入他人的流程,合作很方便。
312+
- **不需要编程**
313+
不会写代码也能用,拉节点、连线就行,谁都能上手。
314+
- **高度自定义**
315+
可以根据自己的需求调整流程,功能很丰富,适合各种场景。
316+
317+
**缺点**
318+
319+
- **界面不统一**
320+
每个人搭建的流程节点可能布局都不一样,看别人的流程可能要花点时间理解和改动。
321+
- **细节太多**
322+
会显示很多底层细节,对新手来说可能有点复杂,有些人只想简单用用会觉得信息量太大。
323+
- **上手有门槛**
324+
虽然不用代码,但理解节点怎么用、流程如何搭比较花时间,新手需要适应。
325+
- **性能受限**
326+
流程复杂或者电脑性能不够时,运行会变慢,尤其是硬件达不到推荐要求的时候。
327+
- **协作有难度**
328+
虽然共享流程很方便,但大家布局风格不同,把不同人的流程拼一起不太容易。
329+
- **自定义带来的负担**
330+
可调的地方太多,有时候会在调细节上花太多时间,影响效率。
331+
302332
## ImportError: cannot import name 'guidedFilter' from 'cv2.ximgproc'
303333
304334
```bash

notes/ai/dev/mcp/serena.md

Lines changed: 17 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -26,4 +26,21 @@ title: serena
2626

2727

2828
```bash
29+
uvx --from git+https://github.com/oraios/serena serena start-mcp-server
30+
31+
uvx --from git+https://github.com/oraios/serena serena project generate-yml
32+
uvx --from git+https://github.com/oraios/serena serena project index
33+
34+
# --transport streamable-http --port 9121
35+
docker run --rm -it --network host \
36+
-v $PWD:/workspaces/projects \
37+
--name serena ghcr.io/oraios/serena:latest serena start-mcp-server --transport stdio
38+
```
39+
40+
- .serena/serena_config.yml
41+
42+
43+
```yaml
44+
# csharp, python, rust, java, typescript, go, cpp, or ruby
45+
language:
2946
```

notes/ai/voice/whisper.md

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -9,6 +9,8 @@ tags:
99
- [openai/whisper](https://github.com/openai/whisper)
1010
- [HN](https://news.ycombinator.com/item?id=32927360)
1111
- https://openai.com/blog/whisper/
12+
- 参考
13+
- https://github.com/Starmel/OpenSuperWhisper
1214

1315
```bash
1416
git clone https://github.com/ggerganov/whisper.cpp

notes/cn/sh/sh-faq.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -11,6 +11,7 @@ tags:
1111
- 积分查询登陆 https://jzzjf.rsj.sh.gov.cn/jzzjf/
1212
- 打印没有二维码时 临时修改站点设置,允许 Insecure content
1313
- 重置后密码为 a111111
14+
- 统一到一网通办了
1415
- 档案核实进度查询 http://222.68.19.232/webMaster/ggrs/ggrs_cxhs.htm
1516
- [紧缺专业目录](http://jzzjf.rsj.sh.gov.cn/jzzjf/pingfen/jqzyml.htm)
1617
- https://etax.chinatax.gov.cn/

notes/db/kv/valkey.md

Lines changed: 19 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -18,3 +18,22 @@ title: valkey
1818
- sponsored and supported by AWS
1919
- 参考
2020
- Wishlist [valkey-io/valkey#17](https://github.com/valkey-io/valkey/issues/17)
21+
22+
# Version
23+
24+
## Valkey 9.0
25+
26+
- HASH field ttl
27+
- HEXPIRE
28+
- HEXPIREAT
29+
- HEXPIRETIME
30+
- HGETEX
31+
- HPERSIST
32+
- HPEXPIRE
33+
- HPEXPIREAT
34+
- HPEXPIRETIME
35+
- HPTTL
36+
- HSETEX
37+
- HTTL
38+
- CLuster 模式支持多 DB
39+
- 原子 SLOT 迁移

notes/dev/dev-awesome.md

Lines changed: 14 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -19,6 +19,20 @@ tags:
1919
- [poteto/hiring-without-whiteboards](https://github.com/poteto/hiring-without-whiteboards)
2020
- http://goo.gl/FsgnGi
2121
Salaries
22+
- Task Runner
23+
- Makefile
24+
- Justfile
25+
- Magefile
26+
- [joerdav/xc](https://github.com/joerdav/xc)
27+
- MIT, Go
28+
- Markdown defined task runner.
29+
- [runme-ai/runme](https://github.com/runme-ai/runme)
30+
- MIT, Go
31+
- DevOps Notebooks Built with Markdown
32+
- [go-task/task](https://github.com/go-task/task)
33+
- MIT, Go
34+
- A task runner / simpler Make alternative
35+
- Taskfile.yml
2236
- Github Copilot
2337
- [CodedotAl/gpt-code-clippy](https://github.com/CodedotAl/gpt-code-clippy)
2438
- formatter

notes/dev/dev-faq.md

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -72,6 +72,14 @@ tags:
7272
- https://www.martinfowler.com/
7373
- https://web.stanford.edu/~ouster/cgi-bin/decisions.php
7474

75+
## Cursed Knowledge
76+
77+
- UGC 域名需要加入到 Public Suffix List - https://publicsuffix.org
78+
- 否则会被标记为恶意域名
79+
- Chrome 使用这个做检测 https://developers.google.com/safe-browsing/reference/URLs.and.Hashing
80+
- https://publicsuffix.org/list/public_suffix_list.dat
81+
- https://immich.app/cursed-knowledge
82+
7583
## 常见命名格式 {#case}
7684

7785
| case | demo | for |

0 commit comments

Comments
 (0)