Vibeleaderboard
Index / tool
Visit github.com
Category
Developer Tools
Rank
No. 2027Tools index
Pricing
Open Source
Type
TOOL
GitHub
1 stars
Latest release
v1.0.0
Date

About

A set of sandboxed, deterministic environments for testing AI agents on tasks set in India, covering trip planning, UPI payments, GST invoicing, form filling, multi-document checks and Hindi and Hinglish customer support. Grading uses written rules rather than an LLM judge, and many tasks are infeasible so agents must say so. It uses simplified, synthetic data and is not legal, tax or financial guidance.

Tags

evaluationbenchmarkai-agentsindiapythontool-use

Tech Stack

Python

Media

Navital Gym

Comments (0)

No comments yet

Editorially curated, with community endorsements as a secondary signal. Corrections welcome.