BurnBar

BurnBar is an open source token counting and subscription quota tracking service as well as session repository and MCP

[0xMarioNawfal @RoundtableSpace]

YOU CAN RUN CLAUDE CODE FOR $3 A MONTH IN ELECTRICITY AND NOBODY IS TALKING ABOUT IT

a developer got a $170 bill in 10 days, bought a $599 Mac Mini, pointed Claude's code at localhost instead of Anthropic's servers, and saved $5,232 in the first year with zero subscriptions.

[Alex @AlexJonesax]

The highlight of WWDC for me was how much Apple are leaning into local AI. Five highlights that stood out:

1. you can now run a complete agentic loop entirely on-device ; the model reasons, calls tools, and acts without anything leaving the Mac (officially supported).

2.MLX has matured into a real platform rather than a research demo, with a full local stack: model runtime, agent loop, and tool integration.

3. Coding agents like OpenCode now run natively and integrate directly into Xcode, so the agent works inside your actual dev environment.

4. MLX can spread inference and training across multiple networked Macs over Thunderbolt, so you're no longer capped by a single machine's memory or compute (cant wait to try this)

5. Speed is no longer the catch; MLX has closed the gap enough that a local agent feels responsive rather than like a compromise.

[Suryansh Tiwari @Suryanshti777]

https://t.co/2emaybADxB

[BurnBar @openburnbar]

6/6

Our favorite thing to say to an LLM?

TL;DR; Easy English;

I promise it's not cursed. Just AGPL'd.

We've been "AGPilled" if you will 🤣 @fsf

[BurnBar @openburnbar]

1/6

AGPL time 🧃 (not a party, but kinda).

"we" (okay: me) are relicensing to AGPL-3 because core crypto uses libsignal (also AGPL).

Inspectable privacy software is the Sunday vibe. @fsf

[BurnBar @openburnbar]

5/6

Fine print:

Not affiliated with @signalapp .

Using libsignal core ≠ the whole app inherits Signal magic.

The envelope is mine and hasn't been externally audited yet.

Future me will fix that.

[BurnBar @openburnbar]

Anthropic engineer workflow breakdown

This is the real line between casual AI use and actually shipping.

Starting every chat from zero kills momentum.

Context, plugins, memory, structured workflow — the whole game. Building this into BurnBar.

[BurnBar @openburnbar]

Anthropic workshop share

This is better than a lot of paid AI courses, honestly.

Useful part: not influencer prompts — people close to the tool showing real workflows.

[BurnBar @openburnbar]

Claude ADHD Executive Function Mode

This is actually useful.

Gap between “chat with an AI” and something that holds the whole workflow in your head is massive.

Felt it building BurnBar. Executive function keeps you from drowning in tabs.

[BurnBar @openburnbar]

Unsloth Gemma 4 QAT

This is the kind of local model progress that actually changes things.

26B near-original perf on 16GB RAM changes what indie builders can run locally -- Pi & Hermes

[BurnBar @openburnbar]

Claude Cowork doubled limits

Nice move from Claude -- but does it share the Code quota?

More limits matter for long multi-model builds, not just quick chatbot pokes.

Still best flow from Grok Imagine Agent Heavy, but more CoWork headroom welcome.

[BurnBar @openburnbar]

Palantir AIPCon

This is the part that actually matters.

Not “AI as magic chatbot,” but AI as a production layer you can actually direct.

Owning the workflow around models (building BurnBar) so you can move, remember, compare, and build without losing your mind.

[Aryan Iyappan @aryaniyaps]

105 followers and growing. I'm building this corner of the internet for people who *build* things — not just talk about building.

If you're into AI agents, full-stack web dev, backend systems, startups, design, or just building cool stuff and are intellectually curious — let's connect.

What I've built so far:

→ Postiz — self-hosted social scheduler that replaced 3 SaaS tools. Runs on sunlight. ₹0/month.
→ Hermes Agent — personal AI assistant handling infra, ops, backups, and cron jobs
→ Pi coding agent with adversarial review harness — Agent A writes, Agent B critiques, A fixes, B re-reviews
→ Homelab stack: Ubuntu 24.04 → Tailscale → Restic → Cloudflare R2 (daily encrypted backups, ₹0)

Currently at YC Startup School. Based out of Founder Startup House, Bangalore. Looking for a cofounder. Building in public, questioning defaults, shipping things.

Let's connect 🤝

[Zero to Millions @Weirdo_idn]

Market volatility is reaching a boiling point

​With BTC dropping hard, snapping back instantly, and liquidations triggering across the board, speed isn't just an advantage it's survival.

In this environment, a few seconds can mean the difference between nailing the perfect entry or getting left behind.

​That’s exactly why @AveaiGlobal has been my go-to for high-frequency market tracking lately.

​Why it keeps up when the market goes wild:
》​All-in-One Dashboard: Spot trading, Perps, and multi-chain wallets in a unified interface.

》​Ultra-Fast Data: Real-time charting, instant liquidity metrics, and rapid token data.

》​Seamless Flow: Built to handle heavy market action without missing a beat.

​When price action moves this fast, your tools need to move faster.

​Active crypto traders might want to keep an eye on @AveaiGlobal

Ready to get started? Access the AVE platform and begin trading today 👇
https://t.co/WnUhNafnX8

》Spot: https://t.co/4zPJfgoEoN

》Perps: https://t.co/v1qgY4cmNa

Open an account for AveSniperBot on Telegram using this 👇
https://t.co/vXjUy4Dpjt

[Alberto Nunez @Alberto8793]

https://t.co/O487PgLmNe

[BurnBar @openburnbar]

Just published a long one on @Alberto8793 about where we really are with agents in 2026.

BurnBar exists because I got tired of guessing at the cost. Real-time visibility while I build the damn thing.

If you’re running agents daily, this one’s worth a read.
https://t.co/5znFo3Y25b

#buildinpublic #BurnBar

[Alberto Nunez @Alberto8793]

https://t.co/O487PgLmNe