https://mnmz81.github.io/loom/en/ Posts and short notes on what I learn — to remember and find again. en https://mnmz81.github.io/loom/en/posts/claude-code-mods-first-mod/ https://mnmz81.github.io/loom/en/posts/claude-code-mods-first-mod/ Mon, 19 Oct 2026 00:00:00 GMT Building a mod that blocks npm commands that skip a wrapper script and adds a /node-guard command: the files, claude plugin validate, a test, running it, letting Claude write it, and the pitfalls. claude-code https://mnmz81.github.io/loom/en/posts/claude-code-mods/ https://mnmz81.github.io/loom/en/posts/claude-code-mods/ Wed, 14 Oct 2026 00:00:00 GMT A mod is a plugin whose JavaScript or TypeScript runs inside Claude Code. How it works, when to pick it over a hook, skill or MCP server, the layers it runs in, and its risks. Version 2.1.295. claude-code https://mnmz81.github.io/loom/en/posts/ai-agentic-rag-graphrag/ https://mnmz81.github.io/loom/en/posts/ai-agentic-rag-graphrag/ Tue, 13 Oct 2026 00:00:00 GMT Two steps beyond a fixed RAG pipeline: letting the model search as a tool, as many times as it needs, and building a knowledge graph for questions that span many documents. ai rag https://mnmz81.github.io/loom/en/posts/ai-agents/ https://mnmz81.github.io/loom/en/posts/ai-agents/ Tue, 13 Oct 2026 00:00:00 GMT What makes a system an agent, how it differs from a workflow, when an agent is worth it and when it is not, a minimal agent in Python, and the Claude Agent SDK. ai agents https://mnmz81.github.io/loom/en/posts/ai-computer-use-skills/ https://mnmz81.github.io/loom/en/posts/ai-computer-use-skills/ Tue, 13 Oct 2026 00:00:00 GMT Agents that operate a desktop through screenshots, mouse and keyboard, and skills: folders of instructions and scripts that give an agent expertise only when it needs it. ai agents https://mnmz81.github.io/loom/en/posts/ai-cost-latency/ https://mnmz81.github.io/loom/en/posts/ai-cost-latency/ Tue, 13 Oct 2026 00:00:00 GMT Where the money and time go in an LLM system, and the levers that cut them: the right model and effort, shorter outputs, prompt caching, the Batch API and streaming. ai llmops https://mnmz81.github.io/loom/en/posts/ai-decision-map/ https://mnmz81.github.io/loom/en/posts/ai-decision-map/ Tue, 13 Oct 2026 00:00:00 GMT The whole series in one place: a decision map for choosing the right technique, what each one changes and costs, a worked example, and a checklist before going to production. ai llmops https://mnmz81.github.io/loom/en/posts/ai-distillation/ https://mnmz81.github.io/loom/en/posts/ai-distillation/ Tue, 13 Oct 2026 00:00:00 GMT What knowledge distillation is, why a teacher's soft probabilities teach more than the right answer, and how distillation of LLMs works in practice: teacher data, filtering, fine-tuning a student. ai fine-tuning https://mnmz81.github.io/loom/en/posts/ai-embeddings/ https://mnmz81.github.io/loom/en/posts/ai-embeddings/ Tue, 13 Oct 2026 00:00:00 GMT What an embedding is, how cosine similarity finds texts that mean the same thing, how to create embeddings with Voyage AI, and where to store them. ai rag https://mnmz81.github.io/loom/en/posts/ai-end-to-end-app/ https://mnmz81.github.io/loom/en/posts/ai-end-to-end-app/ Tue, 13 Oct 2026 00:00:00 GMT One small app built with the whole series: a prompt, RAG over help articles, a tool for live order data, input screening, evals, caching and logging. Each step links to the post that explains it. ai llmops https://mnmz81.github.io/loom/en/posts/ai-evals/ https://mnmz81.github.io/loom/en/posts/ai-evals/ Tue, 13 Oct 2026 00:00:00 GMT Why every LLM system needs an evaluation set, how to build one, code-based checks, LLM-as-judge with a rubric, and how to use evals to compare prompts, models and settings. ai llmops https://mnmz81.github.io/loom/en/posts/ai-fine-tuning/ https://mnmz81.github.io/loom/en/posts/ai-fine-tuning/ Tue, 13 Oct 2026 00:00:00 GMT When fine-tuning beats prompting and RAG, and when it does not, what training data looks like, how LoRA and QLoRA train a model cheaply, and how to check the result. ai fine-tuning https://mnmz81.github.io/loom/en/posts/ai-glossary/ https://mnmz81.github.io/loom/en/posts/ai-glossary/ Tue, 13 Oct 2026 00:00:00 GMT The AI terms from the series in one place, grouped by topic, each with a one-line definition and a link to the post that explains it: tokens, RAG, tools, MCP, agents, LoRA, evals and more. ai llm cheatsheet https://mnmz81.github.io/loom/en/posts/ai-guardrails/ https://mnmz81.github.io/loom/en/posts/ai-guardrails/ Tue, 13 Oct 2026 00:00:00 GMT Layered guardrails for LLM applications: screening input with a small model, validating and redacting output, checking actions, limits, and what to show the user when a check fails. ai llmops security https://mnmz81.github.io/loom/en/posts/ai-how-models-are-built/ https://mnmz81.github.io/loom/en/posts/ai-how-models-are-built/ Tue, 13 Oct 2026 00:00:00 GMT Transformers, attention, pretraining, fine-tuning and preference training (RLHF, DPO): the stages that turn a pile of text into a chat assistant. ai llm https://mnmz81.github.io/loom/en/posts/ai-local-models/ https://mnmz81.github.io/loom/en/posts/ai-local-models/ Tue, 13 Oct 2026 00:00:00 GMT Why run an open-weight model on your own machine, how much memory it needs, what quantization does, and how to call a local model with Ollama, including from the Anthropic SDK. ai llm https://mnmz81.github.io/loom/en/posts/ai-mcp/ https://mnmz81.github.io/loom/en/posts/ai-mcp/ Tue, 13 Oct 2026 00:00:00 GMT What MCP is, why it exists, hosts, clients and servers, tools, resources and prompts, the two transports, and a small MCP server in Python. ai agents mcp https://mnmz81.github.io/loom/en/posts/ai-memory-context/ https://mnmz81.github.io/loom/en/posts/ai-memory-context/ Tue, 13 Oct 2026 00:00:00 GMT How agents keep working when a task outgrows the context window: context engineering, compaction, clearing tool results, notes and the memory tool, with safe file handling. ai agents https://mnmz81.github.io/loom/en/posts/ai-multi-agent/ https://mnmz81.github.io/loom/en/posts/ai-multi-agent/ Tue, 13 Oct 2026 00:00:00 GMT When several agents beat one: the orchestrator-subagent design, what Anthropic measured in its research system, the token cost, and subagents in the Claude Agent SDK. ai agents https://mnmz81.github.io/loom/en/posts/ai-multimodal/ https://mnmz81.github.io/loom/en/posts/ai-multimodal/ Tue, 13 Oct 2026 00:00:00 GMT Sending images and PDFs to a model, how image tokens are counted, what vision is still bad at, and what to do with audio. ai llm