RSSAmplifier

Blog

cohix

Writing here instead of social media. Senior Software Architect at F5. Cabin life, camping, software, dogs, AI, and F1.

blog.cohix.networkRSS feed ↗15 posts

Latest posts

Dynamic workflows with any harness, any model: awman v0.11

I’ve been away from working on awman for a few weeks now. I finished paternity leave and went back to work, so needed a while to settle in! I don’t expect weekly releases every single week going forwards, but the project IS still under very active

A penny for your context?

Welcome back to the PrettySmart show after a week on Vancouver Island exploring what may be the most laid back part of Canada I've witnessed yet: Tofino, British Columbia. This week on the show we have several very different but long-awaited new pieces in  awman : Experimental

Workflows from issue to merged PR: awman v0.9 is out

In the  last post , I announced that  amux  was becoming  awman  — the Agentic Workflow Manager. The rename reflected a shift in focus: the tool has grown from a terminal multiplexer for agents into something closer to a software development lifecycle pipeline tool, orchestrating the

amux is becoming awman: an Agentic Workflow Manager for the entire software development lifecycle

When I started building  amux  earlier this year, the name stood for "agent multiplexer." The idea was simply "tmux, but for agents running in containers", which was immediately useful for me in my daily work. As my "agentic engineering" proficiency grew, the

Behold the miraculous self-improving codebase! ane v0.2 is out.

Since one of the core goals of  ane  is to allow code agents to explore, read, and edit source code with one-shot commands using chords and a simple CLI, this week I experimented with a self-improving codebase to work on  ane  from a code

Introducing ane: A language server-aware chord-based terminal code editor (for humans and agents)

TL;DR ane is a terminal code editor that combines a new chord grammar with language servers to allow both human developers and code agents to efficiently explore and edit source code - ane exec lets agents and scripts make precise reads and edits, while ane ’s TUI editor lets

That refactor sucked, but amux v0.8 is out now

The past few weeks has been spent rebuilding amux's internals from scratch. The previous architecture was mostly “designed” by Claude and Codex across sixty-plus work items, and by the end it had become clear that the three frontends (CLI, TUI, headless) were each implementing business

Code agents are bad at Software Architecture - for now

Hello from paternity leave, week three. There wasn’t an amux release last week because I've gotten so fed up with the patchwork design of the amux codebase I've decided to burn it all to the ground, more or less. Even top-end foundation models

Manage your fleet of agent machines with amux v0.7

Hello and welcome to my first paternity leave blog post! I'm off work hanging with the family for 2 months, but I still plan to continue moving amux forwards while I'm off. With that, one of the reasons I can make so much progress each week

Multi-agent workflows, config commands, and true background yolo mode: amux v0.6

I've been using amux daily for all my personal dev projects the past few months, and though it is working incredibly well for my workflows, there were some codebase architecture issues I was ignoring in order to add all the feature ideas I had bouncing around in my

Code agents that actually get things done while you're asleep: amux v0.5

A few days ago was the first time I was  actually  able to kick off a complicated agent-based workflow before bed and have it completely finished when I got up the next morning. It's really quite impressive what these tools can accomplish these days. Using

More powerful multi-agent workflows: amux v0.4

Running multiple agents in parallel has been pretty addicting as I've been working on  amux , but I wanted to force them to follow my preferred workflow. I wanted to make it repeatable and tweakable over time so I can more closely guide how agents get work done

amux 0.3: Faster specs, better multiplayer visibility

I worked over the weekend to get the core workflow that was introduced in v0.2 feeling smoother: write spec (now even faster), kick off work, validate results, update tests and docs, commit and push. When you are repeating this workflow interleaved across several agents working at the same time,

amux 0.2: Multiplayer code and claw agents in your terminal

As I started working more with code agents and they became entwined in my workflows, I started getting itchy. The itchiness came from running an agent and sitting there while it worked. It felt like a waste, since my brain had already planned out the next 3 tasks I wanted

Introducing aspec: Spec-Driven Development + Containerized Code Agents

After working closely with a number of agentic code tools over the past year, I'm starting to coalesce towards a workflow that has been getting some pretty great results. Today I want to show off  aspec  — a tool that automates this workflow by managing predictable,