# SlackBench: Benchmarking Agents on Collaborative Projects Grounded in Real Code Repositories

> **Platform:** [SPIDITS AI](https://spidits.com/) — Real-Time AI News & Market Intelligence  
> **Published:** 2026-11-05T05:00:00.000Z  
> **Category:** RESEARCH  
> **Impact Score:** 170/100  
> **Primary Source:** [arXiv AI](https://arxiv.org/abs/2611.05512)  
> **Canonical Citation:** [https://spidits.com/timeline/slackbench-benchmarking-agents-on-collaborative-projects-grounded-in-real](https://spidits.com/timeline/slackbench-benchmarking-agents-on-collaborative-projects-grounded-in-real)

## Executive Summary
AI Agents are becoming digital coworkers that can browse the web, write code, run experiments, and assist humans over extended horizons.

## Why It Matters (Strategic Analysis)
SlackBench supplies a concrete, reproducible metric for multi‑agent coordination on authentic development workflows, exposing performance gaps that single‑agent tests miss.

## Referenced Coverage & Sources
- **[arXiv AI](https://arxiv.org/abs/2611.05512)**: SlackBench: Benchmarking Agents on Collaborative Projects Grounded in Real Code Repositories — _AI Agents are becoming digital coworkers that can browse the web, write code, run experiments, and assist humans over extended horizons. Yet existing..._

---
*Synthesized by SPIDITS AI Market Intelligence Desk. Track live AI news, model releases, and funding: [https://spidits.com](https://spidits.com)*