Laying out the streets…STEP 1 OF 3

Gear: Scikit-learn for RSI · Office #54,792

A modular framework for recursive harness evolution.

In the Building AI agents neighborhood (AI and agents). In town since 28 Oct 2026.

Nobody has claimed this office yet. Tablif records every new product and gives it an office; when its owner claims it, the lights go on.

Is this yours? Claim it

What Tablif read on Gear: Scikit-learn for RSI's site

Gear: Scikit-learn for RSI describes itself as: “Gear: Scikit-learn for RSI”

What it is

Calls itself “framework”.

  • Calls itself: framework
“A modular framework for recursive harness evolution.”

What it costs

Has a free plan. Prices quoted: $0.0568, $1.1858.

  • Free plan: yes
  • Price: $0.0568 · $1.1858
“In a separate cost comparison with both models using the same optimized DSH harness ( a0740800 ), Luna max's API cost averaged $0.0568 per task, versus $1.1858 for Astra max — 95.21% lower .”
“Gear is an open-source framework for recursive self-improvement of AI agent harnesses”

Who it sells to

Sells to developers. Available in 2 languages.

  • Sells to: developers
  • Languages: English · 简体中文
“Its modular architecture lets developers implement, compare, and extend optimization algorithms.”
“English | 简体中文”

How it runs

A command-line tool. Distributed through the npm. Open source. AI is the core of the product.

  • Kind: cli
  • Distributed via: npm
  • Open source: yes
  • AI role: core to the product
“Install rsi-gear@latest and agent-hitch@latest using npm and verify that all required dependencies are available.”
“Gear is an open-source framework for recursive self-improvement of AI agent harnesses (prompts, tools, workflows, and context management).”

What it works with

Names Claude Code, Codex and 5 more on its pages.

  • Names: Claude Code · Codex · DSH · +4
“Use Gear through its Refine Skill in Codex, Claude Code, DSH, or another compatible agent environment.”
“Astra + Codex’s 84.1% on AutomationBench’s public Marketing tasks.”

What it offers as proof

Live, no waitlist. Claims “213 stars”. Publishes docs.

  • Number claimed: 213 stars
“You must be signed in to star a repository”
“Stars 213 stars”

Sources

Tablif read these pages, last on 28 Oct 2026. Everything above is what the product says about itself; Tablif does not verify every claim.

Similar products in Tablif Town

Explore Tablif Town