Joined November 2011
- Tweets 2,831
- Following 857
- Followers 75
- Likes 452
20 Photos and videos
Pinned Tweet
Feb 10
What makes OpenDevBrowser different?
- Supports Base64 (Encodes screenshot as binary‑to‑text ASCII characters like A–Z, a–z, 0–9, , /.)
- Supports CLI CDP Extension UI annotation
"npx opendevbrowser"
"npm install -g opendevbrowser"
To start: "npx opendevbrowser serve"
Feb 9
I released an update to opendevbrowser. Give it a try.
github.com/freshtechbro/open…
3
152
May 18
How I setup OpenAI codex agents
In this GitHub gist, I shared my #Codex setup and workflows. Someone in the network might need it.
My Stack: @OpenAI Codex App/CLI, OpenDevBrowser CLI, @RepoPrompt MCP, GPT 5.5 Pro
KISS: Keep It Simple & Small.
gist.github.com/freshtechbro…
1
74
May 5
Similar setup here.
Vibeproxy Codex app
Multiple ChatGPT plus & business subscriptions wired into vibeproxy. Then point codex app/cli to vibeproxy.
My new favorite workflow
=-=-=-=-=-=
Pre-requisites
1. Setup Warp (either clone or use the app)
2. Setup Vibeproxy
3. Use Droid
4. GPT-5.5
=-=-=-=-=-=
Workflow:
- left bar for project management
- files git view
- prooompt GPT-5.5
- Let Droid do your work
1
3
78
Apr 28
OpenDevBrowser v0.0.27
Use any website or component as inspiration for your design.
Opendevbrowser inspects DOM - snapshots - extract design.md - pass advance brief to agents - design pixel perfect custom prototype
npx opendevbrowser
github.com/freshtechbro/open…
1
38
Apr 24
Gpt 5.5 opendevbrowser cli
Point the /inspiredesign tool at any website or component. You get a design.md and a design contract and a prototype.
32
Mar 17
As AI agents and LLM adoption accelerates, lifestyle tasks e.g. shopping will become programmatic & need durable browser automation for agents to shop safely and reliably.
opendevbrowser.dev deliver that. Let your agents shop & hunt for deals on major retailers in stealth
27
Mar 17
Opendevbrowser is solving an important problem for the next generation agent browser automation. The next release is going through release-gate stabilization. If you use AI agents in your workflow, try it. Fully OSS. Thank me later.
opendevbrowser.dev
1
1
35
Mar 7
My take on @OpenAI GPT 5.4: Stay in the 258k limit and you are good. Codex can handle the rest of compaction and keeping you going.
Same old issue: Be very specific with what you want. GPT 5.4 takes instructions literally. Gets most things right at the first try.
1
55
Mar 2
35
Mar 2
OpenPolyTrader for @Polymarket will be released in a moment.
It's a research preview. I am dropping it for those who love the idea of unravelling the prediction markets.
Live data ingestion Multimarket Combinatorial Arbitrage “Frank-Wolfe loop & Bregman projection”
Feb 16
#OpenPolyTrader
A tool I am dropping in the wild soon. Just a bit more tweaks for the lovers of prediction markets. This gives you full control on polymarket arbitrage. Lots of risk management knobs. I have explored the breadth and depth of prediction markets arbitrage and optimized for speed. The UI still needs some love.
This is still in research mode and ensure you check under the hood to see the knobs before you unleash once it's released. I don't plan to host this. It will be fully local and free but if there is sufficient demand, I might host for pennies or you can host online on your personal vps.
A 'clanker" for the polymarket folks to play with.
40
Mar 2
OpenDevBrowser v0.0.16
- CLI Extension CDP @opencode plugin
- Multitab & Multitask
- 10 use cases (QA, Annotate, Data Extraction, Research, Shopping, Deal hunting, UGC asset)
- Extendable (bring your own agent skill)
npm install -g opendevbrowser
opendevbrowser.dev
1
1
78
Mar 1
3D Animation Design skills for AI Agents
I shipped these design skills a while ago. It’s a pack of 3D animation, component animation and motion skills. I will be upgrading the design skills this week with new features, assets and scripts.
github.com/freshtechbro/clau…
5
5
98
Feb 19
Prepping for a new release. A huge upgrade to opendevbrowser. Make your ai agents a first class citizen of the internet.
25
Feb 17
Existing patterns in code plus the fact that the existing code already provides clarity to the llm on how everything works; these make it easy for llms to port code from one language to another.
Feb 16
Yes! I am living in this space right now, and the outcomes are simply wonderful. LLMs are pattern recognizers, and much of the grunt and scut work in coding revolves around patterns.
I’ve modernized a number of apps whose stale UIs were showing their age in ways that stifled usability and functionality.
For example, tackling the extremely difficult challenge of delivering faceted search and browsing over an entity-relationship graph, where relationship (attribute) sizes are unpredictable, creates serious data-scrolling challenges when deployed live on the Web.
x.com/kidehen/status/2019525…
35
Feb 16
This is definitely for the @Polymarket folks.
Github repo up soon. Adding the bells and whistles.
Typescript and @opencode Zen.
Feb 16
#OpenPolyTrader
A tool I am dropping in the wild soon. Just a bit more tweaks for the lovers of prediction markets. This gives you full control on polymarket arbitrage. Lots of risk management knobs. I have explored the breadth and depth of prediction markets arbitrage and optimized for speed. The UI still needs some love.
This is still in research mode and ensure you check under the hood to see the knobs before you unleash once it's released. I don't plan to host this. It will be fully local and free but if there is sufficient demand, I might host for pennies or you can host online on your personal vps.
A 'clanker" for the polymarket folks to play with.
57
Feb 16
#OpenPolyTrader
A tool I am dropping in the wild soon. Just a bit more tweaks for the lovers of prediction markets. This gives you full control on polymarket arbitrage. Lots of risk management knobs. I have explored the breadth and depth of prediction markets arbitrage and optimized for speed. The UI still needs some love.
This is still in research mode and ensure you check under the hood to see the knobs before you unleash once it's released. I don't plan to host this. It will be fully local and free but if there is sufficient demand, I might host for pennies or you can host online on your personal vps.
A 'clanker" for the polymarket folks to play with.
2
1
189
Feb 16
OpenPolyTrader
It's fully autonomous with self healing circuit breakers.
The full release will be loaded with a fully-corrective and adaptive Frank-Wolfe loop with an IP solver and Bregman projection.
Pros: Built-in non-negative combinatorial arbitrage.
Cons: IP solves are expensive/slow, execution-dependent
OpenPolyTrader solves for non-negative edge, latency and cost-efficiency.
Ciao.
15
Feb 11
Found this challenge today and solved it.
total dev time: Approx 5hrs
total runtime: 64s
estimated tokens: 34,465
Actions: 193
Steps: 30
Stack: Opendevbrowser Gpt 5.3
@adcock_brett I hope my submission is not too late.
Feb 2
Solve this in under 5 minutes and I’ll offer you $500k/year in cash plus several million in equity
I'm building a Computer-Use team, goal is to use computers better than humans
No experience or PhD needed
Instructions:
1. Solve all 30 challenges on this website in under 5 minutes: serene-frangipane-7fd25b.net…
2. Feel free to use any tools or vibe code it. Provide us a zip folder with instructions on how to run the agent and reproduce your results, as well your run statistics
3. The agent should be able to solve all the challenges, use browser, and provide overall metrics around time taken, token usage and token cost. Your agent must solve this challenge in under 5 minutes
Email your response: agents@brettadcock.com
If you have any questions about this challenge, feel free to email us
2
1
124
Feb 12
All the artifacts for the challenge solver are on the “feat/browser-automation-challenge” branch in case anyone wants to run it.
github.com/freshtechbro/open…
1
20
Feb 11
Average tokens per session and cache hit rate for the sessions might provide additional context for this comparison. Although i do not expect the ranking to be any different.
But this comparison is only valid if you exclude sessions where users switch between models.
data from 111,000 sessions that were run through opencode.ai/zen
21