acpdbg: let the agent sit at the debugger 🐛
LLM agents are great at static analysis — I keep pasting crash stacks into their context. So why not let the agent read the lldb data itself over ACP, and even drive the debugger? It's a try.
Field notes
Writing
Field notes on AI tooling, written from the deep end as I run into it —
mostly from a 4D developer’s chair. There’s a through-line I didn’t
plan: it opens in March 2025 asking whether anyone had tried MCP, and almost
everything since is the same story from a new angle — the ecosystem quietly
settling on a few boring, shared standards. Spawn a process and speak JSON-RPC
(MCP, ACP, the Copilot SDK, Apple’s fm serve); drop a standard
file at the repo root and let any agent read it (AGENTS.md, Agent
Skills). The loud buzzwords — context engineering, loop engineering —
come and go; the standards are what stick. Newest first.
Pipeline, Flow, or Chain? Picking the Right Tool to Wire LLM Calls Together
Once one prompt's output feeds the next, you have a workflow — and a naming problem, and a tooling choice. When do you reach for a chaining library, and when for a real orchestrator like Airflow or Step Functions?
When Agents Become Planners, What Remains of Workflow Orchestration?
AI agents don't kill workflow engines — they move the DAG down a layer. Agents become the planners; orchestrators like Airflow, Step Functions, and Temporal become the reliable executors.
ACP: Copilot in Xcode, without the plugin 🔌
Xcode 27 speaks the Agent Client Protocol — so you can register the Copilot CLI directly as an agent, pick the model with an env var, and skip the official plugin entirely.
macOS 27's AI dev ecosystem — and pointing 4D AIKit at a local model 🍎
macOS 27 ships the fm CLI (Apple Foundation Models from the terminal), a Python SDK, and Core AI. The kicker: fm serve is an OpenAI-compatible endpoint, so 4D AIKit can drive an on-device model with no API key and no cloud bill.
Loop Engineering — the newest way to burn tokens 🔥
After prompt, context, and harness engineering comes loop engineering — designing a system that keeps guiding and correcting an LLM until the task is actually done.
Spending fewer tokens with Claude Code and GitHub Copilot 🪙
An improvised, battle-tested guide to burning fewer tokens in your IDE assistants — compress terminal output, freeze build scripts, script mass edits, throw away long sessions, slim your CLAUDE.md, clean your MCP servers, and let the AI be terse.
AI and design: a workspace, and a contract 🎨
Two design-with-AI updates landed together. Claude Design is a workspace you design inside; Google's DESIGN.md is a file your agents read to stay on-brand. One's a destination, the other's a contract — and both touch generating 4D forms.
AGENTS.md for 4D — auto-generating a README for your AI agents
AGENTS.md is a README for AI agents: a small markdown file at the repo root that hands an LLM the project's shape so it doesn't burn tokens exploring. Here's one auto-generated for 4D projects — plus the Claude Code import trick.
Reranking in RAG — the precise second pass
Embedding search is a fast first pass that's only roughly right. A reranker is the cheap precision fix: a cross-encoder that re-scores the top candidates by reading query and document together. Here's why it works and what's out there.
Xcode 26.3 as an MCP server — letting agents drive Xcode
Most MCP servers give your agent access to something external. Xcode 26.3 flips it: Xcode exposes itself as a server, so an agent in VS Code (or anywhere) can discover, build, test, and preview a real Xcode project through Xcode's own machinery.
Chrome's built-in AI, and the CEF catch 🌐
Chrome ships on-device AI (Gemini Nano) behind Web-standard APIs — translate, summarize, prompt, all local and private. The asterisk: it leans on Google-internal code that isn't in upstream Chromium, so embedded browsers (CEF) don't get it for free.
Copilot SDK in technical preview — and why a 4D one is within reach
The Copilot SDK gives four languages programmatic access to the Copilot CLI. Under the hood it's the same trick as MCP, LSP, and ACP — spawn the CLI, talk JSON-RPC — which means any app can host an agent.
Agent Skills in VS Code — capabilities that load only when needed
Agent Skills landed in VS Code: folders of instructions and scripts an agent loads on demand. The clever part is a three-level progressive loading scheme, so you can keep dozens installed and pay tokens only for the one that's relevant.
MCP, in practice: GitHub in VS Code — checked in for the team
Eight months after wondering whether anyone had tried MCP, here it is running for real: the GitHub MCP server in VS Code, wired to our issues, projects, and test cases — with the config committed to the repo so the whole team gets it on checkout.
Context engineering — yes it's a buzzword, but this one's real
Context engineering is the step past prompt-writing: assembling everything the model needs — instructions, data, code, memory, tools — in the right format at the right time. A rebrand, sure, but one that points at where the real leverage is.
Apple's Foundation Models — an on-device LLM in a few lines of Swift 🧠
At WWDC 2025 Apple opened the on-device model behind Apple Intelligence to developers via the Foundation Models framework. The API is tiny — and I wrapped it in a CLI and an MCP server to see how far it goes.
Has anyone tried MCP? — and what it could mean for 4D
March 2025: MCP is only a few months old and not yet a standard, but people are already building real things on it. Here's the pitch — and a first sketch of what an MCP server for 4D could look like.
Phi Mage — who's writing, and why the logo is a φ
A short hello: who I am, what the Phimage logo means, and what these field notes will be about — AI tooling, agents, and developer workflows from where the sunlight runs out.