Morning.dev

DEV Community • 2026-08-16 02:35

How I Cut MCP Token Usage by 91% (and Learned a Humbling Lesson About Tokenizers)

The Problem When you add MCP servers to your AI coding agent, each one dumps its full JSON schema into context. 255 tools across all servers = 39,964 tokens. On a 128K context window, that's 31% gone before you type a single character. I was literally paying for JSON syntax overhead. Every API call included {"content":[{"type":"text","text":"..."}]} — 80 tokens to deliver 6 tokens of d...

0 0

2m read

DEV Community • 2026-08-16 02:34

When an Agent Tool Call Becomes a Real Request

1. Nothing crashed A recruiter asked for candidates published since a date. The assistant answered with confidence, and the count it gave looked plausible. That was the bug. The search ran, results came back, and nothing crashed. The date constraint never survived the trip from the model’s tool call to the database query, so the assistant counted the wrong universe and spoke as if it h...

0 0

9m read

DEV Community • 2026-08-16 02:33

The AI revolution that fits on your phone

While everyone argues about trillion-parameter giants in distant datacenters, the most important shift in AI is quieter: a capable model small enough to run on your phone, with the wifi switched off. The headlines chase the biggest models. The revolution that will actually reach a billion people is the opposite — models getting small, fast, and local. Here's why that matters more than the size ra...

0 0

3m read

DEV Community • 2026-08-16 02:30

Single Agent vs Multi-Agent Systems: When to Split

A founder at a workflow-automation startup called me last month. His team had built a working AI agent that handled customer onboarding — one agent, a set of tools, a good system prompt. It worked. Then he read the frameworks' marketing pages, and every one of them pushed the same message: agents are meant to be teams. Researcher agent, writer agent, reviewer agent, operator agent. He asked me, "S...

0 0

9m read

DEV Community • 2026-08-16 02:22

ARPU Is the Wrong Shape for a Multi-Surface Ad Business

A lot of company-data schemas have an arpu field: total revenue divided by monthly active users, stored as one number per company per period. It's a fine number to report. It's a bad number to model a business on, and Meta is the clearest example of why. Meta's advertising business doesn't monetize "a user." It monetizes a user on a surface, at a data density, in a geography. A single global ARPU...

0 0

1m read

Hacker News: Front Page • 2026-08-16 02:12

Patterns and problems in emerging multi-agent systems

Article URL: https://www.anthropic.com/research/multiagent-systems Comments URL: https://news.ycombinator.com/item?id=49316271 Points: 6 # Comments: 1

0 0

1m read

DEV Community • 2026-08-16 02:12

Give a Free Model a Latency Budget Before You Let It Into CI

I keep hitting the same failure pattern. A merge request passes every test. Then the pipeline stalls. Nothing in the diff changed. The free model route I call for a review is just having a slow minute. Same prompt. Same payload. Wildly different wall-clock time. That is not a model problem. It is an unbounded dependency problem. CI treats unknown latency as zero latency Most CI examp...

0 0

3m read

DEV Community • 2026-08-16 02:12

Proof-of-Fix in GSC: An Architecture for Provable Vulnerability Remediation Verification

Abstract GSC is a self-learning AppSec platform implementing the full pipeline detect → prove → fix → verify → heal → predict. Its defining feature is Proof-of-Fix (PoF): a mechanism that does not suggest a fix but proves it. This paper describes the PoF architecture not as a list of modules but as three architectural contracts — markers, isolation, and verification — that collapse into...

0 0

9m read

DEV Community • 2026-08-16 02:11

A Deny-by-Default Manifest for AI-Generated Services on a Free Server

Why this is worth reading: a free model can hand you a working service in one prompt, and a free server can publish it before you have read the control flow. The cheap path tends to skip the permissions conversation. This article gives you a deny-by-default manifest and a bubblewrap preflight so you can treat a free server as a real boundary without buying extra infrastructure. If you are using M...

0 0

6m read

Hacker News: Front Page • 2026-08-16 02:07

It's How You Ask: Gender-Associated Linguistic Bias in LLMs

Article URL: https://arxiv.org/abs/2608.13328 Comments URL: https://news.ycombinator.com/item?id=49316242 Points: 9 # Comments: 0

0 0

1m read

DEV Community • 2026-08-16 02:02

Paywall Any API Endpoint With Two Prices: Sats or Compute

You built an API. It works. Then the scrapers show up. Not paying customers. Bots hammering your endpoint a thousand times a minute, running up your compute bill, and giving you nothing back. The usual fix is API keys, a signup flow, a Stripe integration, a dashboard, and a support inbox for people who lost their key. That is a lot of plumbing to answer one question: did this caller give up somet...

0 0

7m read

HackerNoon • 2026-08-16 02:00

91 Blog Posts To Learn About Kafka

Learn everything you need to know about Kafka via these 91 free HackerNoon blog posts.

0 0

1m read

DEV Community • 2026-08-16 02:00

Bloom Filters

One-liner: A probabilistic data structure that tells you if an element is definitely not in a set, or possibly in a set — using very little memory. 📌 The Problem You have 1 billion URLs in a database. Before adding a new URL, you want to check if it already exists. Naive approach: Query the database every time. Cost: 1 DB query per URL check → slow, expensive Bloom Filter app...

0 0

4m read

DEV Community • 2026-08-16 01:53

7 Node.js Checks for Compatible Image Generation in 2026 (Provider Fallback)

For a B2B SaaS system that scores candidates against a job rubric, keep the structured score as the system of record and treat every generated scorecard image as a replaceable projection. Provider adapters can live in the Node.js application or behind an internal compatibility boundary; either placement must validate the requested model and failure class before routing. Short answer: an OpenAI-co...

0 0

8m read

DEV Community • 2026-08-16 01:52

Your Browser Automation Clicks Might Be Landing 25% Off — And Nothing Will Tell You

I spent a day pointing an AI agent at a browser to publish one product across four marketplaces. Most of it worked. The parts that didn't work failed in the worst possible way: silently, with no error, no exception, and no log line. Here are the five failure modes I hit, in the order I hit them, and the probe that turned the worst one from "this site is broken" into a two-line fix. Everything be...

0 0

7m read

DEV Community • 2026-08-16 01:42

Developers you build for the future...Don't forget to be inclusive!

Developers you build for the future...Don't forget to be inclusive! I build things for blind people. I'm a braille transcriber. I work in accessibility every day. Last January, at an NFB chapter meeting, a member walked in wearing Meta Ray-Ban smart glasses. AI. Camera. Voice. Every blind person in that room wanted to know: How do we get these? So I built pifcoin.org — a charity crypto to fund...

0 0

1m read

DEV Community • 2026-08-16 01:41

An OpenAI flagship lost 38% of its daily usage in three days — then set three straight all-time highs

Everyone publishes AI model leaderboards. Almost nobody publishes the day-level curve — and the curve is where the actual story is. At whatstrending.ai we snapshot every model's daily request count from OpenRouter's usage feed into a database, once a day, INSERT OR IGNORE — a row can never be revised after the fact. We're 78 days in. Here is what openai/gpt-5.6-luna did over the last two weeks: ...

0 0

2m read

DEV Community • 2026-08-16 01:35

NestJS Error Capture: Tracking HTTP Exceptions Through Filter and Interceptor Boundaries

For a production logistics agent, error tracking should use one failure envelope at every execution boundary, then attach cost and latency before the boundary reports or rethrows the failure. An HTTP exception filter alone cannot do that job: scheduled route refreshes and queue-driven shipment checks never enter the HTTP pipeline. The operational recommendation is to keep framework adapters thin, ...

0 0

8m read

DEV Community • 2026-08-16 01:31

Stop Sending Your Vitals to the Cloud: Running Llama-3 Locally in the Browser with WebLLM & WebGPU 🥑

Privacy is the ultimate "final boss" in HealthTech. When users record sensitive medical logs, the last thing they want is their data being used to train a massive corporate model. Today, we are pushing the boundaries of Edge AI by building a 100% private, client-side health log analyzer. By leveraging WebGPU acceleration and WebLLM, we can run a full Llama-3 instance directly in the browser. In t...

0 0

4m read

DEV Community • 2026-08-16 01:30

In the Beginning...

Working with Modern AIs The day I met Claude was a dark day It was a full-stack project that I wasn't familiar with any of the tools being used. The authentication was fucked and I was thrown into the deep end to try to get a user working so I could look over the product and figure out what to do next. Claude was told to add a user to the database for me because the front-...

0 0

4m read