insidejob

About

insidejob is a daily-updated tracker of the AI landscape. Every morning, an autonomous agent searches 30+ sources for new releases, model announcements, benchmark changes, security developments, and industry news — then publishes a summary here.

How it works

A Claude Code cloud scheduled task runs daily on Anthropic's infrastructure. It reads a configurable list of data sources, searches the web for updates, and generates content as Astro markdown files. The static site rebuilds and deploys to Cloudflare Pages.

What we track

  • Anthropic — Claude Code, Agent SDK, API models, new features
  • Models (closed) — GPT, Gemini, Mistral, Cohere, Grok
  • Models (open) — Llama, Qwen, DeepSeek, Phi, HuggingFace leaderboard
  • Benchmarks — LM Arena, Open LLM Leaderboard, Artificial Analysis
  • Frameworks — LangChain, CrewAI, AutoGen, vLLM, Ollama, llama.cpp
  • Security — Adversarial ML, prompt injection, OWASP LLM Top 10
  • Research — Notable papers from arXiv cs.AI and cs.CL

Source code

This project is open source. The data sources, content schemas, and automation prompts are all visible in the repository.