Agent Memory Is Not a Database: Start with an Event Log
I break agent memory into writing, storage, retrieval, assembly, and forgetting, then outline a minimal implementation path that starts with a durable event log.
Search
Find a note.
The Agent Extension Stack: Skills, Plugins, MCP, and Tools
A Skill teaches an agent how to work, a Tool lets it act, MCP connects it to external capabilities, and a Plugin packages the pieces for installation.
Reading World Models Through 'LLMs Can't Jump'
A reading note on abductive reasoning, interactive simulation, and what current world models can and cannot do.
What Pi Coding Agent Keeps in the Core
A source-based comparison of Pi, Codex, and OpenCode shows how a minimal coding harness leaves workflow policy and isolation to the user.
Testing ASR for Video Subtitles
A small Korean-language comparison showed why transcript coverage, timestamps, and manual review matter as much as model output.
Designing a Private Media Processing Workflow
A video-subtitle pipeline that uses Cloudflare for coordination and a local machine for private, CPU-heavy media work.
Inside Matt Pocock's Skills: Where the Control Lives
A commit-pinned reading of the Matt Pocock Skills repository shows how its invocation rules keep workflow direction with the human while the host retains tools and permissions.
A Multi-Signal Design for Follow-Cam Tracking
A follow-cam design that combines face recognition with person tracking, appearance, pose, segmentation review, and hybrid framing.
Tracking One Performer Through a Wide Performance Video
I adapted public ideas from an AI-assisted broadcast workflow into a cautious follow-cam experiment for my bias.
From Superpowers to Skills: What Changed with GPT-5.6
After moving to GPT-5.6, I began choosing smaller agent workflows and paying more attention to the loop that decides what happens next.
How Skills Are Changing the Way I Work with Agents
Reusable skills are helping me give agents clearer working context, spend less time repeating process, and keep the important decisions reviewable.
Moving Two Small Applications to Cloudflare
Field notes on using Workers, Static Assets, D1, R2, Cron Triggers, and Access to move two private applications beyond a local machine.
From NotionNext to Astro: Rebuilding a Personal Website Around Markdown
Why I moved this personal website from a Notion-backed Next.js system to a smaller Astro site with local Markdown content.
From Local RAG Bot to a Cloud-Native Chatbot
How I moved a personal RAG chatbot to Cloud Run with PostgreSQL, session-aware retrieval, and Vertex AI.
From Local Scripts to a Serverless Data Collection Pipeline
How I moved a local Python data collector to Google Cloud so a teammate could trigger it remotely through Telegram.
A Modular Video Translation Pipeline with Whisper
I separated transcription, translation, and subtitle rendering to make a personal video translation pipeline easier to debug.
From Terminal to Browser: Giving My Multi-Agent Chatbot a Face
I wrapped my multi-agent chatbot in Docker and FastAPI so I could use it through a simple browser interface.
Making My Personal AI Assistant Safer and More Reliable
I added safety checks, retrieval ranking, and response editing to make my personal assistant more reliable.
Building a Personal AI Assistant from My Notes
I built a personal assistant around my Obsidian notes to learn how retrieval, agents, and personal knowledge can work together.
A Half-Day Return to Machine Learning with Fraud Detection
A short fraud-detection project helped me revisit class imbalance, Random Forest, SMOTETomek, and precision-recall metrics.
Why I Switched My AI Newsroom from CrewAI to LangGraph
I moved my AI newsroom from CrewAI to LangGraph to gain clearer state, workflow control, and more consistent output.
Building a Weekly AI Digest with an Agentic Workflow
I built a small AI newsroom with CrewAI to collect, filter, and summarize the AI news I actually want to read.
Using Apps Script and Gemini to Triage Gmail
I used Google Apps Script and Gemini to classify incoming Gmail messages and make a crowded inbox easier to scan.
Automating a Mobile App with an Android Emulator
I used Python, an Android emulator, and OCR to automate a repetitive in-app card-spinning workflow.
What LLM Agent Frameworks Actually Add
A practical introduction to agent workflows, using a CrewAI newsroom example to explain roles, loops, and trade-offs.
How to Write Better Prompts for LLMs
Four practical ways to give language models clearer instructions, better context, and smaller tasks.
怎樣才能寫好提示詞？
四個實用方法：講清楚任務、補充背景、要求檢查，再把大任務拆小。
No notes match that search.