AI tool comparison
AI Subroutines vs Browser Harness
Which one should you ship with? Here is the side-by-side panel verdict, pricing read, reviewer split, and community vote comparison.
Automation
AI Subroutines
Record a browser task once, replay it 500x at zero token cost
75%
Panel ship
—
Community
Free
Entry
AI Subroutines from rtrvr.ai are a new automation primitive: you record a browser task once (a form submission, a LinkedIn connection, a CRM update), and that recording becomes a deterministic, callable tool that AI agents can invoke with different parameters — without spending tokens on every run. Unlike Playwright, Browser-Use, or other out-of-process solutions, Subroutines execute entirely inside your browser tab, inheriting your live session cookies, CSRF tokens, and signed headers automatically. The technical approach is clever. During recording, the system captures network requests and DOM interactions, then ranks captured requests to identify the actual API call (filtering out analytics and telemetry). Replay-hostile identifiers are stripped while stable endpoints are preserved. The result is a script that runs in your browser context — no session rebuilding, no key extraction, no proxy rotation needed. The AI handles parameter selection; the script handles execution. The business case is clear for outreach and operations teams: bulk LinkedIn campaigns, CRM mass-updates, scraping pipelines, and form submissions that would cost hundreds of tokens per run instead execute as cheap deterministic scripts. The model positions Subroutines as the "function call" layer beneath AI agents — the actions that don't need intelligence every time they fire.
Browser Automation
Browser Harness
Self-healing browser agent that writes its own missing capabilities mid-task
75%
Panel ship
—
Community
Free
Entry
Browser Harness is a radically minimal Python framework from browser-use that gives LLMs autonomous control over Chrome via the Chrome DevTools Protocol (CDP). The entire codebase is around 592 lines across five files — and that minimalism is intentional. The philosophy: don't constrain the agent with pre-built recipes. Instead, let it identify what's missing and write new domain-skill files on the fly. When the agent hits a capability gap mid-task (say, a tricky CAPTCHA flow or a site with unusual navigation patterns), it authors the missing handler itself and stores it in a domain-skills directory for future runs. Over time, the harness self-improves, accumulating institutional knowledge about specific websites. It also ships with remote browser support — three free concurrent cloud instances — removing the local setup burden. The "Show HN" debut generated early traction for what is fundamentally a different philosophy from frameworks like Playwright or Selenium: instead of comprehensive APIs that try to anticipate every scenario, Browser Harness trusts the LLM to extend itself. This is either the future of browser automation or a maintenance nightmare — probably both.
Reviewer scorecard
“The 'record once, replay many' pattern solves a real cost problem in agent pipelines. The in-browser execution model is clever — you get auth context for free instead of fighting with session management. This is the kind of tool that drops into existing workflows without requiring a rewrite.”
“592 lines of Python is the most impressive part. The self-healing skill-file approach means it gets better the more you use it on a specific site, without any manual intervention. For internal tooling against well-known sites, this is a legitimate alternative to maintaining a brittle Playwright script.”
“Browser automation that runs inside your session is exactly the attack surface that malicious sites exploit. Subroutines executing in-tab with full cookie access means a compromised script could do real damage. The 'zero token cost' claim also obscures that you still need LLM calls for parameter selection — the savings are real but overstated.”
“An agent that writes its own code mid-task is powerful but auditably scary. What exactly is getting written to those domain-skill files? For anything touching auth flows, financial sites, or sensitive data, you want deterministic, reviewable automation — not self-modifying LLM-authored scripts. Pre-alpha warning is warranted.”
“This is the 'compilation' step for agentic workflows — moving from 'LLM decides every click' to 'LLM selects a pre-compiled action.' That separation of concerns (intelligence vs. execution) is how you scale agent operations from one-off demos to production pipelines. The pattern will be widely copied.”
“The principle here — give agents the freedom to extend themselves rather than boxing them into predefined APIs — is the correct long-term direction. Every browser automation framework eventually becomes a sprawling collection of edge-case handlers. Starting from minimal and letting the agent accumulate domain knowledge is cleaner architecture.”
“For creators doing outreach, social posting, or newsletter campaigns, this is genuinely transformative. Recording a campaign action once and letting AI handle personalization at scale is the efficiency unlock that makes solo creator businesses actually viable at volume.”
“For content workflows that involve repetitive browser tasks — scraping competitor sites, pulling analytics, posting to platforms — a self-improving agent that handles edge cases better each time sounds genuinely useful. I'd try it on low-stakes automation first and see how the skill files look.”
Weekly AI Tool Verdicts
Get the next comparison in your inbox
New AI tools ship daily. We compare them before you waste an afternoon.