# uclanlp/OpenVLHarness

> A framework that turns multimodal LLMs into visual agents with nine tools (grounding, OCR, depth, search, Python agent) and multimodal memory, without any fine-tuning.

- Magnitude: 3.7 out of 10 — Early signal
- Stars: 170 total · +5 stars measured 2026-10-22, 00:38–04:50 UTC, ≈ 38 by evening
- Star trust: star growth looks organic
- Category: AI agents · Language: Python · License: MIT · Created: 2026-10-15 · Last push: 2026-10-20
- GitHub: https://github.com/uclanlp/OpenVLHarness · Homepage: https://openvlharness.github.io/ · Page: https://gitnova.dev/en/r/uclanlp/OpenVLHarness

## Useful for

- Run a visual agent on top of Qwen3-VL or GPT to answer image questions
- Build a pipeline with grounding, OCR and depth estimation for spatial scene analysis
- Evaluate a model on 22 benchmarks via the built-in eval suite

## Why it’s here

- Star-counter measurements on 2026-10-22 (UTC), 00:38–04:50: 165 → 170 stars (+5). This is the change over that interval.
- Estimated end-of-day forecast: about +38 stars, using observed gains and the previous day.
- Over the last two days the pace is 5.7× that of the previous 5 days.
- The repository is 7 days old and already has 170 stars. With less than two weeks of history, there's no usual pace to compare the spike against yet.

## Star trust

Star growth looks organic.

Star-trust labels are heuristics based on the repository’s behavior, not a check of every stargazer.

## Numbers

- Forks: 3
- Issues and pull requests: 3
- Watchers: 4
- Average over the last week: 28 per day
- Usual pace: too little history (under two weeks)
- Stars in the last hour (measured): 1

## Stars per day, last 8 days (oldest → newest, today is partial)

2026-10-15 … 2026-10-22: 7, 2, 1, 1, 26, 85, 46, 5

## Magnitude by day, last 2 days

10-21 3.8, 10-22 3.7

## Similar by description

1. **google-research/harness2** — 7.8 · Peaking · AI agents · Python · +150 stars measured 2026-10-22, 00:37–04:49 UTC, ≈ 1,080 by evening
   A framework for recursive optimization of AI agent harnesses (prompts, tools, memory, skills) without fine-tuning model weights, with JobBench, LAB and WorkBuddy-Bench benchmarks.
   Full card: https://gitnova.dev/en/r/google-research/harness2.md
2. **zju3dv/SuperNav** — 0.0 · Steady · AI agents · Python · Not enough measurements today
   An agentic navigation system built on a pretrained multimodal model that solves object-finding and multi-target tasks in Habitat-GS and AI2-THOR without navigation-specific fine-tuning.
   Full card: https://gitnova.dev/en/r/zju3dv/SuperNav.md

---

Magnitude (0–10) measures how fast and how unusually interest in a repository is growing right now. It is not a quality score.
Days are UTC. “So far today” is a fact; “expected by the end of the day” is a forecast.
Summaries and use cases are written by an LLM (DeepSeek V4.1 Flash) from the README and may be inaccurate: verify specific claims (benchmarks, speed, hardware) in the repository itself.
Data as of 2026-10-22 05:00 UTC, updated every 30 minutes.