Crayotter: Traceable Multi-Agent Workflows for Long-Form Video Editing

Sep 11, 2026

Uploaded by Adam

Long-form video editing across heterogeneous footage requires agents to coordinate source selection, multimodal analysis, timeline construction, narration and subtitle alignment, rendering, and revision while exposing intermediate state for inspection and repair. Crayotter is an open-source multimodal multiagent demo system that organizes production around coverage-aware material preparation, artifact-grounded editing research, and tool-grounded timeline execution. Retrieval reports, analyses, blueprints, scheduler events, tool calls, intermediate renders, and final exports are treated as first-class artifacts supporting local assets, monitoring, preview, diagnosis, resumption, and resource-aware asynchronous execution.

LanguageEnglish

CategoryPresentations

FormatPDF

Pages10

Reading Time4 min

AccessFree to download

Abstract

Introduction

Related Work

Crayotter: Traceable Multi-Agent Workflows for Long-Form Video Editing Lecheng Yan 1∗†, Yichong Zhang2∗†, Xiantao Xu3∗, Jianze Lin4∗, Ben Pan5∗, Xiaoyu Zheng6∗, Jiawei Qian6∗, Anqi Wu6∗, Jiahui Geng7 , Ruizhe Li8 , Fengyu Cai9 , Jingcheng Niu9 , Raymond Li 10 , Wenxi Li6 , Chenyang Lyu 11 1University of Science and Technology of China 2Jilin University 3Dmall Inc. 4Beijing Normal University 5Tianjin University 6East China Normal University 7Linköping University 8University of Birmingham 9Technische Universität Darmstadt 10University of British Columbia 11Alibaba Group [lyuchenyang.dcu@gmail.com](lyuchenyang.dcu@gmail.com) arXiv :2606 .07636v2 [ cs .CV] 17 Jul 2026 Abstract Long-form video editing over heterogeneous footage requires agents to coordinate sourceselection, multimodal analysis, timeline construction, narration and subtitle alignment, rendering, and revision while exposing intermediate state for inspection and repair. We present Crayotter, an open-source multimodal multiagent demo system for prompt-driven longform video editing. Crayotter organizes production around coverage-aware material preparation, artifact-grounded editing research, and tool-grounded timeline execution. Across these stages, retrieval reports, video analyses, editing blueprints, scheduler events, tool calls, intermediate renders, and final exports are treated as first-class artifacts rather than hidden transient state. The workbench supports local assets, agent-assisted retrieval, progress monitoring, artifact preview, failure diagnosis, interrupted-job resumption, and resource-aware asynchronous execution for long-running workflows. In a 23-theme evaluation, Crayotter achieves the highest human overall score (3.40/5) among the compared systems, with its largest margins in theme alignment, narrative coherence, and editing smoothness. These results show that long-horizon video editing agents can be made traceable, inspectable, and practically controllable through observable production artifacts. 1 Introduction Long-form video editing transforms a high-level brief and heterogeneous footage into a coherent audiovisual narrative. Unlike short-form generation, it requires joint decisions over material selection, clip understanding, narrative structure, timeline operations, and audio–text alignment. LLM agents enable interleaved reasoning and tool use (Yao * Core Contributor. † Project Lead. ‡Our code is publicly available at [https://github.com/](https://github.com/)[ ](https://github.com/)idwts/Crayotter and our live demo at [https://idwts](https://idwts). [github.io/Crayotter/\#demo](github.io/Crayotter/#demo). et al., 2022) and role-structured collaboration (Wu et al., 2024 ; Hong et al., 2024); visual-production systems extend these capabilities to planning and specialized filmmaking roles (Lin et al., 2023 ; Xu et al., 2025b) . However, they primarily synthesize shots, whereas editing real or retrieved footage requires source-grounded decisions and executable timecodes. Editing agents increasingly support hierarchical media compilation (Li et al., 2026) and longhorizon, music-synchronized execution (Zhao et al., 2026) . Practical systems must also expose why footage was selected, how operations changed the timeline, and where failures occurred. Without such evidence, users cannot inspect intermediate decisions, revise local segments, or resume interrupted runs. We therefore formulate prompt-driven long-form editing as a traceable agent workflow whose production state is observable and reusable. We present Crayotter, an open-source multimodal multi-agent demo system. Unlike pipelines that assume fixed inputs or synthetic assets, Crayotter automatically searches multiple video sources when local material is insufficient. It constructsa coverage-verified material pool, derives a timegrounded blueprint, and executes it through registered timeline tools. Retrieval reports, analyses, plans, tool events, and intermedia

FAQs

What problem does Crayotter address in long-form video editing?

Crayotter targets the coordination challenges of long-form editing, including source selection, multimodal analysis, timeline construction, and narration/subtitle alignment, while making intermediate decisions inspectable and repairable.

How does Crayotter make the editing process traceable?

It treats retrieval reports, video analyses, editing blueprints, scheduler events, tool calls, intermediate renders, and final exports as first-class artifacts rather than hidden transient state.

What capabilities does the Crayotter workbench provide for production control?

The workbench supports local assets, agent-assisted retrieval, progress monitoring, artifact preview, failure diagnosis, interrupted-job resumption, and resource-aware asynchronous execution for long-running workflows.

Sep 11, 2026

Uploaded by Adam