iiinova — where your digital self keeps evolving

Dwayne Forde @dwayne.io · 15/10/2026

Plugins: 80+ internal plugins implement the 2.0 features, TUI plugins are now stable, and plugins hot-reload mid-session.

100 0

Dwayne Forde @dwayne.io · 15/10/2026

The API was rebuilt from scratch, and the agent can call its own API, for example to create a git worktree and move a session into it. Agents setting up their own workspace is a nice direction.

100 0

Dwayne Forde @dwayne.io · 15/10/2026

Background tasks: subagents and shell commands can run in the background, and ctrl+b moves running foreground work there. Useful for long test runs while you keep working.

200 0

Dwayne Forde @dwayne.io · 15/10/2026

Code Mode: the agent writes one short JavaScript script that chains tool calls, instead of separate round trips. MCP servers and @OpenAPISpec endpoints become callable functions. I want to try it on workflows that hit several MCP servers in a row.

200 0

Dwayne Forde @dwayne.io · 15/10/2026

Exciting to see @OpenCode 2.0 ship! is a rewrite, not a point release. Here are the highlights.

100 0

Dwayne Forde @dwayne.io · 15/10/2026

The part I liked most in the claude.ai speedup writeup isn't the 3x, it's the loop: deterministic benchmarks, ship behind a flag, ratchet the benchmark down, repeat. claude.dev/blog/how-we-...

000 0

Dwayne Forde @dwayne.io · 14/10/2026

Gemini 4 Argon is new on the board and comes in sixth, 1.2 points behind the leader Claude Opus 5.5. It is a clear step up from Gemini 3.1 Pro Preview, which scored 92.2% but not enough for a podium finish.

211 0

Dwayne Forde @dwayne.io · 02/10/2026

Claude Code's advisor tool lets a cheaper main model consult a stronger one at key decision points. I'm sure there is a use case to empower some business/operational workflows with this. When coding though, I'll choose to use the best model throughout. code.claude.com/docs/en/advi...

000 0

Dwayne Forde @dwayne.io · 02/10/2026

Latest entrant to the agent execution layer is OpenAI's Dots. The amount of polish and capability from these UXs is getting close a having a fully featured & managed VM. I wonder what will stand out as a key differentiator at this layer.

000 0

Dwayne Forde @dwayne.io · 02/10/2026

Claude Opus 5.5 dropped so we ran it through AWE-bench, and it doesn't disappoint. research.withmantle.com/awe

100 0

Dwayne Forde @dwayne.io · 01/10/2026

All of my most powerful and efficient agentic workflows involved a bunch of agents using an army of CLIs. Really happy @cloudflare.social is leaning into this pattern. 👏 blog.cloudflare.com/cloudflare-c...

000 0

Dwayne Forde @dwayne.io · 01/10/2026

Parenting the age of AI

001 0

Dwayne Forde @dwayne.io · 29/09/2026

New infra layer forming that might upset the IaaS space. Multiple folks vying to own the encrypted, persistent AI memory narrative. deepmind.google/blog/advanci...

110 0

Dwayne Forde @dwayne.io · 28/09/2026

The live interaction experiences with models are advancing so quickly. Some of micro behaviours are a bit unsettling, but compared to where we were last year it's 🤯. Given the option to wait 1hr or use this to check into my hotel, I would probably choose this.

100 0

Dwayne Forde @dwayne.io · 24/09/2026

All the frontier AI models got a bump this month. Seems to be a pattern forming which means the next major bump will be in Dec/Jan. I can't think of another time where the tech stack evolved this quickly. research.withmantle.com/llm-costs/

000 0

Dwayne Forde @dwayne.io · 04/09/2026

Been a while! dAppwright is up to v2.13.12 🎉 - MetaMask 13 support, custom networks included - Coinbase Wallet bumped to 3.143.0 - Headless mode now works natively without xvfb - Wallet projects can run in parallel 🚀 - Test chain moved from ganache to hardhat github.com/TenKeyLabs/dappwright

000 0

Dwayne Forde @dwayne.io · 03/09/2026

More new models from @anthropic.com but what's interesting is the shifting benchmarks towards deeper knowledge work. As more and more models become proficient at generating code, it might not continue to be the key differentiator for useful model inference. www.anthropic.com/claude-fable...

000 0

Dwayne Forde @dwayne.io · 03/09/2026

After spending some time with cmux, herdr, and @zed.dev threads over the last couple of weeks. I've settled into Zed as my tool of choice. It combines the value of an AI TUI with worktree support and a full featured editor. So I can quickly traverse and review all the branches of generated code.

130 0

Dwayne Forde @dwayne.io · 26/08/2026

The big unlock for me was having great vim support, keeping something that can interface with the rest of the team (they are on VSCode), and figuring out that the Zed Agents pane allows you to open up a terminal so I can run native CLIs.

100 0

Dwayne Forde @dwayne.io · 20/08/2026

Hackathons have always been part of the default rhythm at @Mantle. Skills and tools are evolving at record pace, and teams need the space to explore/evolve freely. We set a day, a theme, and let the creativity fly. 🚀

011 0

Dwayne Forde @dwayne.io · 19/08/2026

The industry has been busy! Interested to see if the model releases ramp up again in Q4. 📊

000 0

Dwayne Forde @dwayne.io · 18/08/2026

Long overdue update for the LLM pricing calculator. There has been a lot of new models and the tool checks daily for new pricing as vendors post updates. llm-calculator.withmantle.com

100 0

Dwayne Forde @dwayne.io · 17/08/2026

The message "Now I can iterate without you." is something I strive for with all of my agentic workloads. Sometimes requires giving access to some third party tooling but totally worth it.

000 0

Dwayne Forde @dwayne.io · 13/08/2026

When burning tokens locally, make sure you're using you're charging with as many watts as possible. My OWC dock charges at 80 watts, it was not enough as my laptop was ranging from 100 - 150 watts of power draw under load.

120 0

Dwayne Forde @dwayne.io · 15/06/2026

The Co-Founder Connect panel discussion was genuinely rich. The perspectives shared by Jasmine Cumberland and Derek Hopfner were incredibly insightful, and the audience engagement really made the experience meaningful! ✨

110 0

Dwayne Forde @dwayne.io · 27/04/2026

Daniel de Repentigny & James Ma had standout talks on MCP orchestration and facilitating dynamic AI workloads on Cloud Run. The kind of sessions where you walk away rethinking your own architecture. 👏

110 0

Dwayne Forde @dwayne.io · 27/04/2026

Just got back from Google Cloud Next '26 in Las Vegas. The message was clear: agentic AI is no longer a pilot. It's in production. The whole industry is moving. A few things that stood out 🧵

100 0

Dwayne Forde @dwayne.io · 06/03/2026

Heading to Orlando for #InvestOps2026 next week and joining a great panel on how investment ops leaders can cut through the noise and pick tech that actually moves the needle. Come say hi or grab drinks → luma.com/uwsjm8wjIf

010 0

Dwayne Forde @dwayne.io · 24/02/2026

Someone updated the intro to Silicon Valley 🤩

050 0

Dwayne Forde @dwayne.io · 17/02/2026

My laptop's workload for the past couple of months now that agents are constantly tinkering on things in the background.

100 0

Dwayne Forde @dwayne.io · 03/02/2026

Thanks to UTMIST for inviting me to Friday's panel alongside founders Anthony Azrak and Rachel Wong.

111 0

Dwayne Forde @dwayne.io · 26/12/2025

dAppwright v2.13.0 released with @metamask.io 13 support. 🚀 Delayed support was due to being able to control the backup and sync functionality which really slowed everything down at scale, and to adopt support for the new sidepanel ui. github.com/TenKeyLabs/d...

110 0

Dwayne Forde @dwayne.io · 23/12/2025

For years, I've had a spreadsheet that consolidates current pricing for major models to track and compare the cost of different AI workflows across the company.

100 0

Dwayne Forde @dwayne.io · 23/12/2025

In the last couple of weeks, 3 new models dropped and I caught myself scrambling to keep up with the latest pricing.

100 0

Dwayne Forde @dwayne.io · 06/10/2025

Wonderful night hosted by YSpace, York University celebrating some amazingly ambitious startups with a solid lineup of pitches. Nafis Ahmed mentioned "A community is only as strong as each of its nodes" to set the tone for the night.

110 0

Dwayne Forde @dwayne.io · 04/10/2025

Effective context management unlocking the next wave of efficiency with agents. I've talked to many organizations where their journey starts off strong and then they start to suffer from "context rot" and their effectiveness starts to trend downwards.

110 0

Dwayne Forde @dwayne.io · 02/10/2025

This report recognizes 100 technology leaders shaping the future of our industry, through community nominations. I've had the pleasure of speaking at one of their flagship events, and was instantly bought into the quality and efficiency of the organization as a whole. 👏

110 0

Dwayne Forde @dwayne.io · 02/10/2025

I’m truly honoured to be included and featured in the #CTOCraft 100 alongside so many inspiring technology leaders. A special thank you to the #CTOCraft community for the recognition, and for creating a space where leaders can learn, share, and grow together. 🙌

131 0

Dwayne Forde @dwayne.io · 26/09/2025

Great all the AI vendors are sharing their market data around AI effectiveness. At events, it feels all the CTOs are trying to compare notes and tooling and baselining their efficiency with each other. Reports like this aren't necessarily complete, but really help the conversation.

110 0

Dwayne Forde @dwayne.io · 13/09/2025

Last night I was involved in a very insightful panel hosted by Marcia Dunbar and team over at the Onyx Initiative. We had a deep conversation around how #AI is affecting the job market, our work habits, and what employers are looking for in the next generation of workforce.

000 0

Dwayne Forde @dwayne.io · 08/09/2025

We've been exploring the VSCode AI statistics feature (code.visualstudio.com/updates/v1_1...) recently. Wondering where this number will land over a long period of time and over all the different approaches (refactoring, test generation, prototyping, etc) a healthy codebase needs.

020 0

Dwayne Forde @dwayne.io · 29/08/2025

Great image showing show the progression of AI agents within engineering teams. I find that our practices are all over this chart, so it could benefit from an y-axis with engineering task complexity. Still, really liking how this progression is articulated.

010 0

Dwayne Forde @dwayne.io · 21/08/2025

Had a blast speaking on building a successful business in Canada at Founders Connect in Toronto. My three takeaways: - Know your market so well it feels like a co-founder. - Iteration speed wins — ship, learn, and improve fast. - Use Canada’s advantages — especially SR&ED

031 0

Dwayne Forde @dwayne.io · 13/08/2025

There’s something special about building a company in Canada. 🇨🇦 This week at Founders Connect Connect in Toronto, I’ll be sharing what it’s taken to build here: what’s worked, what’s surprised me, and why I believe this is one of the most exciting times to start something meaningful.

010 0

Dwayne Forde @dwayne.io · 07/07/2025

Have #AI experiences won the right to have the first the interaction purely due to a guaranteed sense of safety/support regardless of the question? 🤔

020 0

Dwayne Forde @dwayne.io · 04/07/2025

Toronto Tech Week was a whirlwind — a much-needed hard reboot for the city’s tech scene. It was energizing to see so many notable people come out to fuel the buzz, support one another, and embrace this exciting wave of technological change.

121 0

Dwayne Forde @dwayne.io · 19/06/2025

Had an incredible time at Google #Gemini Day for Startups in Toronto! The energy was electric! Demoing and collaborating on ideas around tracewright.com, our open-source regression test automation agent, was icing on the cake. 🤩

030 0

Dwayne Forde @dwayne.io · 29/05/2025

I had an amazing time at the CTO Craft Con Toronto event. So many insights and lessons from industry leaders that are grounded in their real world efforts which makes all the information instantly relatable and tangible. Big congrats to the whole CTO Craft team on a flawless event.

041 0

Dwayne Forde @dwayne.io · 13/05/2025

Sometimes when filling out a form it decides to fill out one field at a time, reads the form errors, then continues. Prompt: - Switch to the "Test Company" Corporation - Create a new Share Class with only the name "Voting Common", prefix of "VC". Leave the other fields blank

100 0