
Agents From Anywhere - How to Set Up Mobile-First Agent Orchestration for Any Codebase

nader dabit
@dabit3
2d
Agents From Anywhere - How to Set Up Mobile-First Agent Orchestration
This setup gives you:
• Sandboxed environments for every session
• Parallelism across many tasks
• Prompt agents from your phone, Slack, Linear, web app + more
• Agents push PRs directly when a task is https://t.co/ja60R0DyIq
0
1
7
Understanding the agent stack might be the most valuable technical skill you can build right now.
Pi is one of the most interesting projects I've seen in a while and has everything you need to build custom agents.
I broke it down here, How to Build a Custom Agent Framework with PI:
1
1
13
Life update - I'm joining Cognition!
More here:
10
2
50
You can battle and compare up to 5 models including Kimi K2.5 and Opus 4.5 out of the box with Arena Mode from Windsurf.
Humans are way better at spotting differences between two outputs than judging one in isolation, so I think this will become a very popular workflow going forward.
x.com/dabit3/statu......
1
1
13
New Video - How to Build an Effective Long Running Agent Loop in 7 minutes.
This video walks you through the entire process from creating a spec, building and polishing a PRD, to running the agent.
1
0
21
You can literally vibe code your own beautiful niche personal portfolio in 15 minutes with tools like Windsurf or Claude Code.
In 2026, there is no good reason not to have your own!
Here are the prompts for 5 different terminal portfolios I built, along with the code and links for you to try them out yourself:

nader dabit
@dabit3
1mo
You can literally vibe code your own beautiful niche personal portfolio in 15 minutes with tools like @windsurf or Claude Code.
In 2026, there is no good reason not to have your own!
Here are the prompts for 5 different terminal portfolios I built, along with the code and https://t.co/3DxQUK3WRe
1
2
22
EigenDA hit 1 GB/s throughput on mainnet. ⚡️
Enough to handle 1 trillion rollup transactions per day, scaling ETH blobs 10,000x.
And the architecture is built to 10x from here, with a goal of reaching TB of throughput.
Powered by @ethereum
EigenDA hit 1GB/s.
Sustained, cloud-scale throughput across 100 operators globally — not a demo, not a shortcut.
Built via:
• LittDB (custom DA storage, ~1000× faster than LevelDB)
• GPU-accelerated encoding (20× latency drop w/ Ingonyama)
• Fully modular data & control planes
For verifiable systems, data availabili
2
6
25
New Video - Deploying any application or agent to a TEE in 4 minutes ⚡️
• Give smart contract guarantees to any app or agent
• No infrastructure management - deploy like Vercel
• Persistent private keys across reboots and upgrades
• TEE runtime and memory encryption - no one accesses your data
• Observable upgrades wi...
7
24
102
TEEs explained in 150 seconds ⚡️
TEEs let you build private and verifiable applications with the performance + flexibility of web2.
• Private prediction markets with verifiable resolution
• AI agents with their own wallets.
• Private order books.
• High-performance games.
• Privacy-preserving applications that can'...
7
9
66
Me and my team are headed to Buenos Aires for DevConnect this week. 🇦🇷
I'm excited to meet as many new people IRL as possible, so if you're in town check out our events below and come hang out!
2
6
29
Deterministic inference: getting the exact same output every time you run an LLM with identical inputs.
Try it yourself:
0
6
25
