gm legends โจ
hope your Tuesday is starting off smoothly. Grab a cup of coffee and tuck in for another fine edition of the Leaderboard. In today's issue: Anthropic's new coding sidekick, Zapier is stepping into the agentic era, and a more thoughtful habit tracker. Let's dive in.
Anthropic does it again

Claude 3.7 Sonnet is Anthropic's latest AI model, introducing a hybrid reasoning approach that offers both rapid responses and detailed, step-by-step problem-solving. This flexibility allows users to choose between quick answers or in-depth analysis, depending on their needs. Alongside this, Anthropic has unveiled Claude Code, an agentic coding tool designed to assist developers directly from their terminals.
๐ฅ Our take: Everyone in AI seems to be taking a different approach to coding. OpenAI has ChatGPT as a generalist, xAI is chasing a more independent problem solver, Bolt is a do-it-all app generator, and Cursor is focused on AI-first workflows. Anthropic seems to be aiming for something more groundedโAI that fits into how developers already work instead of trying to change the process entirely.
Automate all the things

Zapier Agents takes automation beyond simple triggers, letting you create AI-powered assistants that handle complex tasks across 7,000+ apps. These agents can access live data, process information, and take initiative on commands, making automation feel more like an assistant than a static workflow.
๐ฅ Our take: AI automation sounds great until you spend more time fixing it than it saves. Zapier Agents promises to handle tasks without the usual micromanagement, which could be a game-changer if it actually delivers. If itโs smart enough to work without constant tweaking, it might finally make AI automation feel effortless
Your agents stopped answering and started doing
Your agents call tools, write to databases and move money, so a wrong answer is now a wrong action, which is a considerably worse meeting. Arcjet runs policy before every LLM call, tool call, query and API request: prompt injection, sensitive data, bots, rate limits, spend quotas, and your own rules in Rego on top. It reads the OpenTelemetry you already emit, so switching it on is one environment variable, no code changes and nothing to deploy. Each check also reads the earlier steps in the same workflow, so an agent's tenth action gets judged against the first nine and not on its own. JavaScript, Python and Go SDKs all at 1.0, agent framework guards included, free tier to start on.
A more thoughtful habit tracker

HabitGo is an app designed to help you build better habits for a happier, more productive life. It assists in planning daily schedules and tracking emotions to foster healthy routines and maintain balance. By encouraging repeated positive actions, HabitGo aims to provide new perspectives and boost confidence. Whether you're looking to establish a morning routine, exercise regularly, or practice mindfulness, this tool offers a structured approach to achieving your goals.
๐ฅ Our take: We've all downloaded habit-tracking apps with the best intentions, only to abandon them after a week. HabitGo seems to understand this common pitfall by integrating emotion tracking with habit formation, recognizing that our feelings play a significant role in sticking to new routines. It's like having a personal coach who not only reminds you to hit the gym but also checks in on how you're feeling about it.ย
Hot off the presses ๐ฅ
Robots for the home, startup shakeups, and criminally underrated tools. Here's what's hot on Product Hunt Forums today.
๐ค Would you let a robot in your home or not?
๐จโ๐ป What's your favorite criminally-underrated AI coding tool?
๐ค Has anyone tried Clay for managing relationships and networking?









