Log inSign up
Evil Martians profile banner
user avatar

Evil Martians

@evilmartians
Design and engineering consultancy for developer tools, AI, and cybersecurity startups. Lago, Tines, Teleport, Wallarm, and Whop grow with Evil Martians.
New York, NY
evilmartians.com
Joined August 2008
177
Following
9,725
Followers
RepliesArticlesMedia
  • Pinned
    user avatar
    Evil Martians
    @evilmartians
    Oct 15, 2025
    Agentic traces contain perfect information about an agent’s behavior with every plan, action, and retry. But that information gets lost in a sea of JSON. So we built AgentPrism: open source React components that turn traces into visual diagrams for debugging AI agents. You can
    Image
    00:00
  • user avatar
    Evil Martians
    @evilmartians
    7h
    We know Rails and AI go well together, but we didn't have official data to support this claim. We do now. This dream collaboration with the Rails Foundation will have two stages, both part of the same benchmark. First, we analyze agent performance on atomic tasks and evals.
    user avatar
    Ruby on Rails
    @rails
    13h
    Today we shared the first results of Agents on Rails, a new, ongoing initiative to measure how well today's leading agentic coding tools (both frontier and open-weight) actually perform on Ruby on Rails codebases. Built by @evilmartians, commissioned by the Rails Foundation.
    Agents on Rails. Benchmarking how well today's leading agentic coding tools perform on Rails codebases. Built by the team at Evil Martians. Commissioned by the Rails Foundation.
  • user avatar
    Evil Martians
    @evilmartians
    11h
    Rails + AI models, which to use? Claude Opus 4.5 is slightly more accurate and less expensive than Fable, although a tad slower. However, you can get essentially the same results with Kimi K3 for almost half the price. We know this because we worked with @rails to benchmark the
    Image
  • user avatar
    Evil Martians
    @evilmartians
    Aug 12
    After an AI startup found us via Claude, we measured relevant traffic for two months. AI was more than double: 268,000 agent requests & 107,000 human views. Site audiences are increasingly machines, so now we publish directly for them, too. Check the new post in the comment from
    Image