Fetching from the wire…
01
02
03
04
05
06
07
08
Markets2026-09-13 · source-backed
Apify stopped accepting new rental Actors on April 1 and fully retires the model October 1, after which unmigrated Actors move to pay-per-usage and their developers earn nothing. The replacement has developers define billable events such as producing a result, uploading a file or starting a run. The stated reason is that event billing suits agent consumption better than a flat fee. This is a marketplace deleting subscriptions outright rather than adding a usage tier next to them, which is a stronger signal than any pricing-page A/B test.
Each link below shares sources, entities, or timing with this story.
One year ago today, Andrej Karpathy fired off what he called a "shower thought throwaway tweet" and accidentally named an entire industry. Vibe coding. The numbers since then are staggering: 92% of US developers have adopted vibe coding practices, the AI coding market hit $8.5...
Between September 3 and 4: Glama's TDQS scoring registry, a zero-execution MCP Security Auditor on Apify, Agent Looker on Product Hunt auditing risky agent behaviors, and ARBR, an open-source framework for oversight of API calls in AI applications. This is a different layer fr...
James Kettle published the whitepaper August 5, presented at Black Hat and DEF CON (PortSwigger). The architecture detail is what agent builders should copy: Turbo Intruder got an MCP interface plus Python orchestration, and the exploitation agent's script was deliberately spl...
DSB-IFEval tests the gap between how voice agents get benchmarked and how they get deployed. Benchmarks hand the model explicit turn-management instructions; production agents get a role description and have to infer when to listen, backchannel, interrupt or yield. Across 1,03...
Neon Innovation Lab's auditor uses pure static analysis on the syntax tree and configs, deliberately avoiding execution because running untrusted MCP code is itself an RCE path. It detects command injection (CWE-78), path traversal (CWE-22), secret exposure (CWE-798), unauthen...
Google launched agentic video understanding across Gemini 3.7 Flash, 3.6 Flash and 3.5 Flash-Lite (Google). Instead of scanning a video start to finish at a fixed sample rate, the model runs an internal loop deciding what to watch, at what speed, and through which channel: fra...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.