coles.codes

herdr vs cmux, after two months on each

Sun, 02 Aug 2026 00:00:00 +0000

I left cmux for herdr at the start of June because its remote support did not fit how I work. Maybe I should have waited a little longer! cmux shipped a daemon that keeps SSH terminals alive across reconnects on 1 June, then added SSH agent forwarding on 4 June.

I stayed on herdr anyway. I’ve now spent about two months with each, moving between Linux boxes and my Mac while running several coding agents at once.

Reviewing code you didn't write

Wed, 22 Jul 2026 00:00:00 +0000

Most of my last couple of weeks have been spent on code reviews and planning. Everyone can write code fast and cheaply, teammates and agents both, so the wall of PRs mounts quicker than anyone can read it. Reviewing is how I stay a contributor to all that code: reading a change properly, even just its core path, means I know what’s going on in that part of the system and I can still challenge the approach with everything I know about the rest of it. Skip the review and I’m not really involved in that code any more, just adjacent to it.

Where I run my coding agents, and why it depends on the box

Sun, 19 Jul 2026 00:00:00 +0000

I use Pi and Claude Code day to day, and I run them in two places. On my Mac, Claude Code’s permissions stay on because the box holds personal credentials and uncommitted work. On a pair of second-hand Dell micros, both agents run unattended with scoped project access and no irreplaceable local state.

I assume an unattended agent will eventually do something I didn’t mean, so the useful safety question is what that mistake can reach. Hooks and instruction files still help inside the session, but the machine and its credentials decide how much damage is possible.

Designing MCP tools for agents

Fri, 17 Jul 2026 00:00:00 +0000

The tempting way to build a data MCP server is one query_data(dataset, filters) tool. Every new slice adds another branch to its filter grammar.

The agent clients I use give the model a code execution environment, so I first ask whether the server needs to shape the result. If the caller may see the full dataset, the tool can return the rows and let the agent work with them in Python or DuckDB. A small skill can explain the dataset’s conventions.

Why local models get stuck repeating JSON

Tue, 14 Jul 2026 00:00:00 +0000

I disqualified a 30B vision model for being too slow, and I was wrong about why. On my DGX Spark it timed out past 21 minutes on trivial documents, with no reading gain over its 8B sibling to show for the wait, so I binned it. Then I re-ran the same task on an RTX 5090 and successful runs finished in 22 to 140 seconds, so speed was never the real problem. The model still fails, because under grammar-constrained decoding it drops into a repetition loop that no sampler setting I tried gets it out of, and on the Spark that loop had just run slowly enough to look like a timeout.

Pydantic AI structured outputs and evals on Bedrock

Mon, 13 Jul 2026 00:00:00 +0000

Building and securing MCP servers with FastMCP

Fri, 10 Jul 2026 00:00:00 +0000

Most MCP tutorials stop at a few tools running on your own machine with no authentication. That is enough to learn the protocol because you are the only person who can call them. Company data already has access controls around it, so a tool that answers anyone who asks bypasses those controls.

Skills or MCP servers: when you need a server

Wed, 08 Jul 2026 00:00:00 +0000

The argument I keep hearing is that skills have made MCP servers redundant. A skill can drive any CLI, so point the agent at aws or psql and it can do anything a server can. For one person working on their own machine, that’s mostly true.

Calling a CLI doesn’t make a skill a good way to authenticate or retrieve governed data. The skill uses whatever credentials happen to be on your laptop, which means the caller is whoever you last logged in as. It can do anything those credentials allow. There is no central place deciding which tools finance can see or recording who touched a customer’s data. The skill may also give the agent more access than the task needs.

Anchoring specs to code with ast-grep

Mon, 06 Jul 2026 00:00:00 +0000

Specs rot for a reason: we link them to code by file path and line number, and code moves. The spec says the token refresh logic lives in src/auth/service.py, someone refactors it into TokenService, and now the spec describes a file that no longer exists. Nobody notices until the spec is wrong enough to mislead someone, usually an agent, usually mine.

Instead of pointing at where the code lives, match what it looks like. Anchor the spec section to an ast-grep rule for a function called refresh_token inside TokenService, and moving the file no longer breaks the link. A path is an address; the rule recognises the code’s shape. Here’s how each survives the same edits:

Breaking In

Thu, 18 Jun 2026 00:00:00 +0000

If you’re thinking about getting into software engineering right now, the vibes are terrible.

Every second post on social media tells you the career is dead: AI writes the code and nobody hires juniors, so pack it up.

I understand why people feel that way, and some of it is true, but the full picture is messier than the doom scroll suggests and a lot more hopeful. I’ve been building software for over a decade, these days as a principal engineer with AI tools open every working day, so here’s what I think you need to know before you write the whole thing off.

The shapeshifting engineer

Wed, 17 Jun 2026 00:00:00 +0000

This post is for the mids, seniors, and principals trying to work out what our jobs look like now that the machine writes the code. If you’re early in your career and wondering whether the whole thing is still worth pursuing, that’s a different conversation and I’ll write it up separately soon.

I’ve been building software for over a decade, and these days I use AI coding tools every working day. The velocity is real. I’ve shipped features in an afternoon that would have taken me most of a week, and cleared backlog items that sat untouched for a year because the effort never justified the payoff.

Local models in mid-2026: the engineering that closed the gap

Fri, 12 Jun 2026 00:00:00 +1000

Herding parallel agents on a remote box with herdr

Wed, 10 Jun 2026 00:00:00 +0000

My personal coding setup is a remote Ubuntu box that I connect to via ssh. Pi and Claude Code do the work. I scope down access because Pi runs without permission prompts and I use Claude’s --dangerously-skip-permissions flag. Neovim handles edits, while yazi gives me review and filesystem navigation with git status patched into the file tree. tmux and mosh hold it all together so nothing dies when my connection drops, and it’s been good to me. The Mac in front of me is basically a thin client unless I’m working on something UI-heavy.

Building lgtmaybe: a PR reviewer for any model

Sun, 07 Jun 2026 00:00:00 +0000

I built a PR reviewer called lgtmaybe, which is the joke I wanted in the name before I’d even started writing the code. You point it at a pull request, pick a model with one flag, and it posts inline comments plus a summary. A clean PR gets a 👍 LGTM, and everything else gets a maybe.

The thing I’m happiest with is that you can run it on OpenAI, Anthropic, OpenRouter, Bedrock, Vertex, or a local Ollama box.

About

Sat, 30 May 2026 00:00:00 +0000

The morning coffee, my first priority!

Hey, I’m Matt - a Principal Engineer at AWS, living in Melbourne, Australia 🇦🇺. I build things and write up the bits worth keeping.

Lately, that’s mostly:

Building tools in Python - mostly Pydantic, PydanticAI, FastAPI and AWS Strands. I like typed code, and agents do a lot of the boring work.
Playing with open-source LLMs - Qwen, GLM and friends running locally on a Framework Desktop and a DGX Spark. Lately I’ve been testing OCR and vision models for fun.
Running a homelab - Raspberry Pis, an old Synology NAS, Dell OptiPlex Micros and the bigger AI boxes, all joined by Tailscale and running lots of containers under Docker Swarm and Portainer.
Living in the terminal - Claude Code and Opus daily, Ghostty on the Mac with herdr inside it, and endless tinkering with my Linux and Claude Code configs.
Making apps - Into Flutter lately, after plenty of Swift and React Native. Backends usually built with CDK on AWS.
Talking about it - I speak at user groups and conferences like AWS re:Invent and PyCon AU, make the odd YouTube video, and used to host a stream called Devs in the Shed.

At work I ship software and worry about how it’ll behave under real load. My backends usually start as FastAPI modular monoliths, with services split out only when something genuinely needs to scale on its own.

Friends

Sat, 30 May 2026 00:00:00 +0000

A few blogs and people worth your time.

Paul Kukiel - a good friend of mine, and one of the best Solution Architects I know. He writes hands-on posts on AWS, Python, and AI-assisted coding.

Melchi Salins - a colleague who goes deep on compute and AI infrastructure. His KV cache explainer even ships with an interactive calculator.

If we know each other and you’d like a link here, get in touch.

Projects

Sat, 30 May 2026 00:00:00 +0000

A few things I’ve built or am still tinkering with.

lgtmaybe #

A PR reviewer that works with six model providers behind one --provider flag. Bedrock and Vertex use GitHub OIDC, so the cloud providers don’t need static keys in your secrets. The same Python core ships as a PyPI CLI and a GitHub Action, with AI doing most of the typing. I wrote up how I built it, the full docs live at lgtmaybe.coles.codes, and the source is MattJColes/lgtmaybe.

Identifiers within AWS CDK

Thu, 30 Sep 2021 00:00:00 +0000

Archive post - originally published 30 September 2021 on devsintheshed.com. The example is CDK v1 in JavaScript, and I’ve transcribed the code from the original article’s screenshots (the two lines of requires and the class declaration above the visible area are the standard cdk init template). The concept of construct IDs and unique IDs is unchanged in v2, but the imports have moved and there’s a renaming trap worth knowing about, both covered in the 2026 appendix at the end.

Reference and import existing assets into AWS CDK

Thu, 30 Sep 2021 00:00:00 +0000

Archive post - originally published 30 September 2021 on devsintheshed.com. I’ve left the article exactly as it ran back then: CDK v1, aws_cdk.core, the per-service aws-cdk.aws-* packages, the lot. It’s a snapshot of how this worked at the time. CDK has since moved to v2 and a fair bit of the code below no longer runs as written - skip to the 2026 appendix at the end for what’s different now.

Hey everyone,