Video yükleniyor...

Video Yüklenemedi

Ana Sayfaya Dön

Very nice Claude Code dashboard for power users to track every request. - Use multiple subscriptions with load balance to switch accounts - Set models for subagents - Track latency, token usage, and costs in real-time - Full request logging and error traces (Link below!)

110,838 görüntüleme • 1 yıl önce •via X (Twitter)

38 Yorum

Nifal Adam profil fotoğrafı
Nifal Adam1 yıl önce

Only for API? A load balance for subscription usage would have been awesome

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

Supports Pro and Max plans too

Nifal Adam profil fotoğrafı
Nifal Adam1 yıl önce

Damn. Got to check it out. Right now my usage is under $200 but interesting

Raymond Weitekamp profil fotoğrafı
Raymond Weitekamp1 yıl önce

hey @kieranklaassen this made me think of you. i think @danshipper will approve at least two max subs per @every employee ;)

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

@kieranklaassen @danshipper @every I’ll take an extra max sub too plz

Jess.py profil fotoğrafı
Jess.py1 yıl önce

this looks cool but youre doing this for learning purpose & the views right? because you will get alot of people pissed , considering we all received the email that says we'll be rate limit soon

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

It’s not mine, I found it on Reddit and liked the project. I think the rate limits will be okay and hopefully will improve how people “vibe code” to be more conservative with tokens!

Kurishita profil fotoğrafı
Kurishita1 yıl önce

Mind is slightly different. But I can directly start and stop agent tasks and can add new objectives to an agent directly from my dashboard.

Alex 🇵🇹 profil fotoğrafı
Alex 🇵🇹1 yıl önce

Much needed after last announcement from Anthropic 😅

zeo profil fotoğrafı
zeo1 yıl önce

AI builders: ToS? never heard of it

Karim C profil fotoğrafı
Karim C1 yıl önce

this is exactly what mario needed when we hit rate limits at 3am the multi-subscription load balancing is chef's kiss. we've been manually swapping API keys like animals real question: does it handle the anthropic vs openai context window differences automatically? mario processes restaurant analytics with varying document sizes and the context juggling gets messy production AI needs this level of observability. most people think AI deployment is just "call API" until their first 429 error at scale

Michael Giannulis profil fotoğrafı
Michael Giannulis1 yıl önce

Super cool how do you fund this stuff so quick

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

Reddit usually! Or lots of people email me their projects

Michael Giannulis profil fotoğrafı
Michael Giannulis1 yıl önce

You keep me busy haha

Harun R. profil fotoğrafı
Harun R.1 yıl önce

Wow this is what power users need

Sunny profil fotoğrafı
Sunny1 yıl önce

with this you can calculate your usage for that weekly limit?

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

@sunnypause I don’t think so unless Anthropic are transparent with the exact limits

L profil fotoğrafı
L1 yıl önce

Nice!!!

Kevin Salzmann profil fotoğrafı
Kevin Salzmann1 yıl önce

Does it also work for pro users?

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

Yup, Pro and Max plans supported

AllAboutAI profil fotoğrafı
AllAboutAI1 yıl önce

super handy

BEN SYNE profil fotoğrafı
BEN SYNE1 yıl önce

made by @snipeship by the way 📈

Kevin Schawinski profil fotoğrafı
Kevin Schawinski1 yıl önce

lol very good

Rob Schmidt profil fotoğrafı
Rob Schmidt1 yıl önce

Have you tried out Devin AI yet? How does it compare to CC?

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

I haven’t but they emailed me a bunch of times to try it. I will try and test it out soon!

Jack profil fotoğrafı
Jack1 yıl önce

Really nice find. The load balancer is a great idea

Faheem | FrontierMind AI profil fotoğrafı
Faheem | FrontierMind AI1 yıl önce

This dashboard looks super helpful for tracking usage across multiple Claude accounts. Load balancing between subscriptions and real-time monitoring of latency and costs would be great for managing budgets and debugging. Thanks for sharing!

Ariel Frischer profil fotoğrafı
Ariel Frischer1 yıl önce

This is cool - I also want to share which essentially for claude code power users who want to automate their workflows and run multiple agents in the cloud.

TreasureHunter profil fotoğrafı
TreasureHunter1 yıl önce

Does it work with bedrock as well

Ali Sherief profil fotoğrafı
Ali Sherief1 yıl önce

It's great but I don't think I have any utility for this as a Pro user.

MangoMagic™️ profil fotoğrafı
MangoMagic™️1 yıl önce

Track requests like a hawk. Transform data into your secret advantage.

Ashish Sheth profil fotoğrafı
Ashish Sheth1 yıl önce

Did I see $35 spend in last one hour? Is that correct? Definitely we need this tool.

YT profil fotoğrafı
YT1 yıl önce

is that possible to assign specific model for different sub agents through prompting?

Ian Nuttall profil fotoğrafı
Ian Nuttall1 yıl önce

I don’t think so yet!

Malcolm Little • Anatrova profil fotoğrafı
Malcolm Little • Anatrova1 yıl önce

I have a question for making a dashboard for a mobile app. Should I code it inside of the mobile app or create a whole new code base file that is a web app if that makes sense.

James Blackwell profil fotoğrafı
James Blackwell1 yıl önce

Noice

Prasenjit Sarkar profil fotoğrafı
Prasenjit Sarkar1 yıl önce

How about this collection of sub agents?

الحلم البعيد profil fotoğrafı
الحلم البعيد1 yıl önce

@grok "شرح هذا المنشور بالكامل وتوضيح المقصود منه بأسلوب مبسط ودقيق."

Benzer Videolar

how to use claude code mods like a top 1% user, step by step: give this to your agent before everyone catches on👇 1. set up Jev connect Jev to the model registry you want to use. check that the connection works and the listed models are available. 2. build your claude code mod open claude code 2.1.287 or later and paste this prompt: “build a mod called run-ledger. load plugin-authoring and use the API types for my installed version. create a dashboard that shows: - estimated cost per run, model, and source plugin where known - which model handles each task - each subagent’s status, latest action, and elapsed time include token counts, cache usage, and reported retries. count each request once. keep background tasks linked to their original run. use dated prices. label costs as API estimates, not subscription charges. show unknown when data is missing. connect the mod to my existing Jev setup. give Jev the task, available models, prices, budget, and relevant past results. ask it to recommend a model and explain why. start with recommendations. make automatic routing optional for eligible subagents. show the recommended model, actual model, result, and cost. include Jev’s own cost. keep state across hot reloads. add details and export. the dashboard itself must make no model calls. validate the plugin. test rendering, costs, attribution, and duplicate counting. give me steps for a live test.” 3. test the setup allow hot reload when prompted. run a simple task, a subagent task, and a mod-triggered model call. check that the dashboard updates, each request counts once, and Jev’s recommended model can actually run the task. 4. install the working mod ask claude to copy it out of the temporary folder and install it as a persistent plugin. see the cost. choose the model. check the result.

Avid

47,276 görüntüleme • 24 gün önce