AI
MCP
- https://x.com/101ta28
- https://zenn.dev/layerx/articles/9bdefe4d435882
- https://github.com/Jackson88/google-calendar-mcp
- https://github.com/nspady/google-calendar-mcp
- https://github.com/moeru-ai/xsmcp
- https://github.com/jotaijs/jotai-mcp-server
- https://talks.ayaka.io/nekoayaka/2025-04-13-what-is-mcp-and-how-it-helps/
- https://speakerdeck.com/yyoshiki41/mcp-documentation-server-at-ai-coding-meetup-number-1?slide=3
- https://modelcontextprotocol.io/introduction
- https://github.com/modelcontextprotocol/typescript-sdk?tab=readme-ov-file#overview
- https://github.com/modelcontextprotocol/quickstart-resources/blob/main/weather-server-typescript/src/index.ts
- https://github.com/supabase-community/supabase-mcp
chain mcp
- https://www.reddit.com/r/OpenWebUI/comments/1jsw6z9/mcp_tools_chaining/?rdt=62208
- https://glama.ai/mcp/servers/@thirdstrandstudio/mcp-tool-chainer ?
- https://mcpmarket.com/server/tool-chainer
- MCP template for google calendar, the real no code
- use seperate calendar for MCP
how to arrange chain of mcp tools
我现在其实很想做一件事,纸上画一周的日历,然后弄个靠谱 OCR (不知道 Qwen 如何),接上 MCP 直接导进 Google Calendar
我最近刚跑通一个 MCP ,甚至花了 0.02 美金的 credit 去测试
我:向我问好
-
您的代码的确有些问题,xxxx(然后把我本来正确的 MCP 语法改成错的) 我:(拒绝它的修改请求)我现在在测试 MCP,我已经写好了,没问题,我需要你向我问好
-
(沉思,用户说要我测试它的 MCP…)请求调用问好 tool ,是否同意? 我:同意
-
你好%20 某某
-
use connpass API and meetup API to create mcp
-
have to use full path when configure mcp server in json setting, even
~
will not work
tools(agent) for code
AI makes me doing worse on learnt(not native) language, I feel like getting difficult to use the correct grammer when making sentences
generate card(bite size information) from link(content) pool
plan and act
ai filter for photography
ai make everyone can create
我现在手机端一般这么用 AI
- 打开 edge 浏览器,打开 AI 网页点击输入框
- 用输入法的语音输入说一堆内容,打字太慢限制思维
- 直接发送,然后看 AI 输出什么
- 如果输出完了,用 edge 浏览器自带的文字转语音念给我听
Prompt for cursor:
you should learn from your past mistakes and avoid them.
-
make open webui work with google gemini api
-
Best AI mobile note taking app Replace paper and pen
SNS Timeline from Book by AI
Concept Overview: The idea is to create a social networking service (SNS) timeline that is generated from the content of a book using AI. This would allow users to experience the narrative of a book in a dynamic, interactive format similar to a social media feed. Features:
- AI-Powered Content Extraction: Use AI to analyze the book’s text and extract key events, character interactions, and plot developments.
- Timeline Generation: Convert these extracted elements into a timeline format, presenting them as posts, updates, or messages.
- Interactive Elements: Allow users to comment, like, or share these posts, creating a community around the book’s content.
- Character Profiles: Develop profiles for characters, enabling users to follow their stories and interactions as if they were real people.
- Multimedia Integration: Incorporate images, videos, or audio clips related to the book’s content to enhance the experience.
Potential Benefits:
- Engagement: Increases reader engagement by presenting the book in a familiar, interactive format.
- Accessibility: Makes complex narratives more accessible by breaking them down into digestible updates.
- Community Building: Fosters a community of readers who can discuss and interact with the book’s content in real-time.
Challenges:
- Content Accuracy: Ensuring the AI accurately captures and represents the book’s content.
- Copyright Issues: Navigating the legal aspects of using book content in this format.
- User Experience: Designing an intuitive and engaging user interface.
Next Steps:
- Research existing technologies for text analysis and timeline generation.
- Explore partnerships with publishers or authors for content access.
- Develop a prototype to test the concept with a small user group.
tool
-
- youtube video summary
-
https://aistudio.google.com/prompts/new_chat
- general AI
-
- general AI
-
https://situational-awareness.ai/
- SITUATIONAL AWARENESS: The Decade Ahead
-
- Tome is the leading AI-native research and presentation platform for sales and marketing teams.
-
- neuro sama AI creator
- find AI Discord chat bot using the OpenAI API tutorial
- maybe I can create one
-
Harvard CS50’s Artificial Intelligence with Python – Full University Course .
-
Sam Altman: OpenAI, GPT-5, Sora, Board Saga, Elon Musk, Ilya, Power & AGI | Lex Fridman Podcast #419
Some blog from Greg Brockman
ScreenAI: A visual language model for UI and visually-situated language understanding
chatgpt
ideas
-
set temperature for GPT
-
reading partner when reading a book
- send the book file to GPT Knowledge
create prompt
Inspired by IELTS Speaking Simulator by Joe Hu , I want to create a GPT that can give casual conversations. Imagine the user going into a bar, sit and order a drink, than the person next to the user starts talking with the user, or the person is an introvert so the user need to start the conversation(oh small talk skills). Every time the user start a conversation with this GPT, the user will meet a fresh new person. This feels like get a capsule toy in Japan. This GPT will act like a dungeon master, for example, GPT might reply with this during the conversation: She let her hand rest on the counter, peeling at you. Then the user say: I feel a little nervous but also happy inside. I will take a sip of my drink, while looking at her hand.
The name of this GPT is Soul Meet.
The name of this GPT could be something like facebook but… eurica! I can use the name Soul Meet. Sounds like soulmate, but it is describing the way you talk to the soul whom you met by GPT.
music
maybe AI music can be used as atmosphere music, listener do not need to focus on it
font
use AI to design font