HN Showcase
Back to browse
GitHub Repository
3 starsTypeScript

Sentinel – browser agent using 3x+ fewer tokens (open benchmark)

by isoldex·May 19, 2026·1 point·1 comment
Visit ProjectView on HN

AI Analysis

●●SolidBig BrainNiche Gem

Self-benchmark shows Sentinel uses 57x fewer tokens than browser-use on hard tasks.

Strengths
  • •Raw JSON data committed for independent verification of the methodology.
  • •Tests reliability across multi-step flows where other frameworks often fail.
  • •Highlights architectural inefficiencies in competitors re-reading page DOM.
Weaknesses
  • •Self-authored benchmark lacks third-party validation of the results.
  • •Only compares three specific frameworks against a single model variant.
Category
AI/ML
Target Audience

Developers building AI browser automation agents

Similar To

BrowserUse · Stagehand · LangChain

Similar Projects

Developer Tools●●Solid

MAKO – Open protocol for LLM-optimized web content (93% fewer tokens)

MAKO compresses what matters into a HEAD-friendly payload — frontmatter, declared actions and semantic links — so agents can find relevance without downloading 181KB of navigation, ads and scripts. The project ships a spec plus real tooling (typed SDK, Express middleware, an analyzer/score and edge-friendly /md conversion), which is a rare combo of protocol thinking and usable developer ergonomics. Whether it becomes a standard depends on buy-in from CMS/plugin authors and agent platforms, but technically it's a smart, practical swing at an obvious pain point.

Big BrainSlick
juanisidoro
113mo ago