Klura turns one-off Claude browser tasks into reusable MCP capabilities and replays them up to 1,400× faster
rundfunk · reddit · 2026-07-24
Klura is a source-available MCP runtime for Claude that tries to turn one-off browser tasks into reusable capabilities.
- The problem: browser agents often rediscover the same website flow every time, wasting time and tokens.
- On first execution, Claude completes the task in the browser while Klura captures browser state and network activity.
- Klura then reverse-engineers the workflow into a reusable capability, preferably as a direct HTTP request with variable parts exposed as arguments.
- If direct requests are not possible, it stores a page script that runs in the site’s own JavaScript context.
- The capability can fall back to a recorded browser path for edge cases like login renewal, and can be used across devices with Claude Code + Dispatch.
- Benchmarks claim replay is 50×–1,400× faster than raw browser automation; one ASOS task went from 1m 35s to 67.9ms.
- The author is looking for feedback from people building with MCPs or browser agents.
More from coding & agent
- 'My intelligence stack': developer splits AI coding work across Opus 5.5, Fable and Astra with human review gates — sh_reya · 2026-10-27
- Baseten webinar: evaluate open models for coding agents by benchmarking on your own merged PRs — baseten · 2026-10-27
- Thorsten Ball on working with agents: run more evals, automate everything, use the best models — omarsar0 · 2026-10-27
- DeepMind paper: LLMs can't sample randomly on their own — pass random numbers from code instead — rohanpaul_ai · 2026-10-27
- BELL: Open-Source Stack Combines Browserbase, Eve, iMessage and Stripe Link for Personal Agents — jeff_weinstein · 2026-10-27
- Legora runs 2,000 parallel agent evals on Modal Sandboxes for legal AI — blelbach · 2026-10-27