The Most Intuitive AI Coding Agent - Code faster, smarter, and stay in the flow.
Joined May 2024
- Tweets 537
- Following 297
- Followers 1,291
- Likes 170
167 Photos and videos
Pinned Tweet
May 18
Codex. Claude. Gemini.
Pick one? Or run all three. ⚔️
3
5
31
617,839
Jun 12
Same task: add dark mode to a website. Three models, three bills.
All three shipped working code. Opus 4.8 was fastest. M3 was ~3x slower but cost half as much. GPT-5.5 ran ~3x the price of M3.
Fastest is not always cheapest. Watch the run.
80
Jun 12
Everyone can build an agent now.
Securing it and shipping it to production is where the real work starts. Spec it, review every diff, keep it inside your repo's context.
Build is the demo. Ship is the job.
1
40
Jun 11
One task: add dark mode to a website. Three models, three very different bills.
MiniMax M3 · $0.768
Opus 4.8 · $1.31
GPT-5.5 · $2.24
All three shipped working code. M3 came in at ~40% the cost of GPT-5.5 (slower, but it got there).
Pick the model for the job, not the hype.
1
101
Jun 11
Code review is quietly becoming the highest-leverage hour in an engineer's week.
Not because there's less code. Because there's 10x more of it, and most of it was written by an agent.
Who's reviewing the reviewers?
1
41
Jun 10
Renaming one function across 80 files used to be an afternoon. Now it's one clear instruction.
The leverage was never in writing new code. It's in safely changing the code you already have.
1
47
Jun 9
The senior engineering skill in 2026 isn't writing code.
It's writing a spec precise enough that the agent can't get it wrong. The bottleneck moved from typing to thinking.
1
4
87
Jun 8
An agent that hasn't read your codebase is just a confident intern.
Context is the whole game. An agent that knows your architecture and your past decisions writes code that actually fits.
3
53
Jun 5
The filename isn't the bottleneck. Coverage is.
23 of the 40 most popular repos on GitHub have no agent file at all. agents.md vs claude.md does nothing for a repo that has neither.
The real win is an agent that reads the codebase and builds that context itself.
Jun 5
Everyone's fighting over agents.md vs claude.md.
I scored the 40 most popular repos on GitHub. 57% have no agent instructions file at all. Not agents.md, not claude.md, nothing.
React, Django, Rust, Vite, Supabase, none of them.
We're debating which filename wins. Most repos don't even have one.
1
123
Jun 5
Autocomplete makes one developer type faster.
Shipping software is a team sport. That gap is the whole difference between a coding assistant and an engineering platform.
51
Jun 4
The fear is "AI replaces developers." The reality is sharper: it exposes the gap between typing and thinking. James breaks down the one skill that separates the two 👇
Jun 4
Everyone's scared AI will replace developers.
After 20 years writing code, I think they've got it backwards.
AI won't replace you. It will expose you.
Here's what I mean 🧵
1
95
Jun 4
Claude Code vs Codex vs Gemini. Which model for what. Save this one: a cheat sheet for routing the right model to the right task.
103
Jun 4
Claude plans the work. Codex writes the code. Gemini reviews the diff.
One orchestrator keeps all three in sync, in your repo, with full context.
That's not a prompt. That's a pipeline.
64
Jun 3
Uber now caps AI coding tools at $1,500 per engineer, per month.
That's $18,000 a year, per developer.
A budget like that didn't exist two years ago. The tools earned it.
78
Jun 3
AI-generated code ships with ~1.7x more bugs. The fix isn't writing less AI code. It's a second model reviewing the first. Different models fail differently, so the bugs one misses, the other catches. We saw 34% more issues caught this way across 200 PRs.
2
61
Jun 1
A full e-sim app, iOS and Android, shipped with the Zencoder extension.
One developer, the whole codebase in context, the agent handling the cross-platform plumbing.
May 23
using zencoder extension (Claude) to dev. e-sim, breaching the gap between local and international MTN or Airtel network, bypassing country's geo restrictions. Available in iOS and Android
1
205
Jun 1
The smartest AI tool is a high bar to claim.
What backs it: the agent reads your whole codebase before writing a line, then routes each step to the model that handles it best.
Apr 29
zencoder has made it the smartest ai tool.
1
107
Jun 1
Goes beyond coding is the whole point.
The agent reads the entire codebase, plans the work, then routes each step to the right model. Writing the code is the easy part now.
May 4
ICYMI: Zencoder goes beyond coding thenewstack.io/zencoder-goes…
1
71
Jun 1
Claude, Codex, Gemini, and Grok usually fight over your codebase.
We put them on one team and let them play the final. 🏆
100% AI-generated. Built with Zenflow.
1
2
4
55,339
Jun 1
The make-of: ~$60, 2 days, zero cameras.
Higgsfield Seedance 2.0 Nano Banana 2 for the shots, @elevenlabs for sound, Remotion for the comp.
@higgsfield_ai this one's for you.
2
1
173
Jun 1
The point under the metaphor: one model plans, one builds, one reviews, one orchestrates.
That's multi-model coding with Zenflow. The best AI for coding isn't a model, it's the routing.
Watch the full trailer: youtu.be/WE_wvH9zMhs
1
45