this account is no longer active. follow me at @techfrenAJ
Joined July 2025
- Tweets79
- Following38
- Followers198
- Likes187
AJ @ Augment Code retweeted
Support for Kimi K2.6 is now here in Augment Code!
We teamed up with @Kimi_Moonshot and @FireworksAI_HQ so Augment users get the latest open-weights frontier model.
This is our first open-weights model and we'd love to hear what you think. Try it out now!
Meet Kimi K2.6: Advancing Open-Source Coding
🔹Open-source SOTA on HLE w/ tools (54.0), SWE-Bench Pro (58.6), SWE-bench Multilingual (76.7), BrowseComp (83.2), Toolathlon (50.0), Charxiv w/ python(86.7), Math Vision w/ python (93.2)
What's new:
🔹Long-horizon coding - 4,000+ tool calls, over 12 hours of continuous execution, with generalization across languages (Rust, Go, Python) and tasks (frontend, devops, perf optimization).
🔹Motion-rich frontend - Videos in hero sections, WebGL shaders, GSAP + Framer Motion, Three.js 3D.
🔹Agent Swarms, elevated - 300 parallel sub-agents × 4,000 steps per run (up from K2.5's 100 / 1,500). One prompt, 100+ files.
🔹Proactive Agents - K2.6 model powers OpenClaw, Hermes Agent, etc for 24/7 autonomous ops.
🔹Claw Groups (research preview) - bring your own agents, command your friends', bots & humans in the loop.
-
K2.6 is now live on kimi.com in chat mode and agent mode.
For production-grade coding, pair K2.6 with Kimi Code: kimi.com/code
-
🔗 API: platform.moonshot.ai
🔗 Tech blog: kimi.com/blog/kimi-k2-6
🔗 Weights & code: huggingface.co/moonshotai/Ki…
How does a Principal Engineer at @Adobe orchestrate AI coding agents? Going LIVE with Lars Trieloff to find out this Friday 10am PT
RSVP: luma.com/psg5f9xf
Going LIVE this Friday with @Wattenberger to see how a Lead Designer actually builds in the post-IDE era
She's the mind behind Intent. This should be wild.
AJ @ Augment Code retweeted
The best way to get proficient at building agentic workflows? Watch over the shoulder of someone building with AI.
We're starting a livestream series where we pair with senior engineers to see how they build.
In our next session, @sambreed (ex‑Senior Staff SWE @ Credit Karma, now Partner @ Sutter Hill and Intent engineering lead) will walk through how he’s rebuilt his own development workflows around spec‑driven development, including:
· How he structures specs before writing code
· How and why he uses worktrees
· How he thinks about ADE vs IDE vs CLI, and when he chooses each
Register here: luma.com/b3u2al1c?utm_source…
AJ @ Augment Code retweeted
Intent is our vision for what comes after the IDE.
AI has changed how we build software. But, it’s also made our workflows messier.
One agent is great. Two work. Past that, things fall apart fast. Prompts go stale, context lives everywhere, and you end up spending more time on the tedious work of orchestrating agents.
The bottleneck isn’t writing code anymore. It’s keeping the agents aligned.
That’s why we built Intent.
The worlds best semantic search is now FREE to try with ANY agent!
It improved the performance of @cursor_ai and @claudeai code over 70%!
AJ @ Augment Code retweeted
Yes, there are a few differences:
* The code review quality is higher thanks to a customized agent we built for this purpose. Our agent scores highest in quality of all code review agents we've tested on a Greptile-based code review benchmark.
* Extensive customizability of code review guidelines.
* Analytics dashboard showing engagement, performance and ROI
* Ease of installation (3 clicks to install)
AJ @ Augment Code retweeted
@augmentcode takes #1 with highest recall on @greptile Code Review Benchmark
Introducing Augment Code Review, powered by GPT 5.2. It's the #1-ranked AI code reviewer across precision, recall, and overall quality.
Free for the first week for every paying customer, and free for open source projects.
Finally you will be able to use @augmentcode industry leading context engine in any Agent!!!
We want to bring @AugmentCode Context Engine to more agents and applications.
I coded up an MCP server that works in Claude Code, Cursor, @zeddotdev, @GitHubCopilot, @kilocode and more. Makes agents more codebase-aware and smarter.
Early version, but give it a try! You need an Augment account. The MCP server won’t consume any credits for now.
docs.augmentcode.com/context…
AJ @ Augment Code retweeted
We just added Prompt Completions to Augment for VS Code.
You’ll now get context-aware suggestions as you type prompts for your agent — including natural-language completions, variable names, file references, and more.
It makes writing great prompts faster, especially for users who craft long specs or detailed instructions.
Powered by full codebase context and agent history, so agent prompt completions feel in sync with what users are trying to write.
AJ @ Augment Code retweeted
Excited for Augment to be featured at OpenAI Dev Day! If your codebase is large and messy (who’s isn’t??), you should try augmentcode.com
AJ @ Augment Code retweeted
I’m at OpenAI DevDay! Hit me up if you’re here and want to chat about agents/code/..
AJ @ Augment Code retweeted
It's Auggie CLI Hackday over at @augmentcode, what better time to get Auggie in @zeddotdev.
AJ @ Augment Code retweeted
Claude Sonnet 4.5 is now the default model for Augment Code.
We’re rolling it out to all customers over the next 24 hours, where it will be available alongside Sonnet 4 (for a limited time) and GPT-5 in the model picker.
Here’s how Sonnet 4.5 compares to Sonnet 4:
AJ @ Augment Code retweeted
OpenAI’s new GPT-5 Codex is impressive at coding tasks. But, in a huge codebase, intelligence isn’t the main bottleneck—context is.
We tested in Chromium (250K+ files, 60GB).
Prompt: “Find code that handles user file uploads.”
Codex: keyword searches, a few file reads… and pointed to the wrong places.
What you really need: indexing + context-driven search that maps the full upload flow (dialogs → security → validation → submission).
Takeaway: raw model intelligence is useful, but without deep codebase context, it falls short in real-world systems.
AJ @ Augment Code retweeted
Augment Code just introduced a $20 plan with 125 messages per month… Good price, but is 125 enough?
Probably not! But it might be enough for people to get hooked.
