# CodeCRDT: Observation-Driven Coordination for Multi-Agent LLM Code Generation

arXiv: https://arxiv.org/abs/2510.18893 · HTML: https://mlsanity.com/paper/2510.18893/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use.
- Builds on 4 of its 17 references that are in the map.
- First posted on arXiv 18 Oct 2025.

## Placed in

- LLM agents and tool use / [Software engineering coding agents](https://mlsanity.com/agents/software-engineering-coding-agents/)

## Builds on (most cited first, 4 of 4)

- [MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework](https://mlsanity.com/paper/2308.00352/) (2023)
- [ChatDev: Communicative Agents for Software Development](https://mlsanity.com/paper/2307.07924/) (2023)
- [The Impact of AI on Developer Productivity: Evidence from GitHub Copilot](https://mlsanity.com/paper/2302.06590/) (2023)
- [Collaborative Text Editing with Eg-walker: Better, Faster, Smaller](https://mlsanity.com/paper/2409.14252/) (2024)

## Built on by (most cited first, 11 of 11)

- [Verified Detection and Prevention of Concurrency Anomalies in Multi-Agent Large Language Model Systems](https://mlsanity.com/paper/2606.17182/) (2026)
- [Multi-agent Collaboration with State Management](https://mlsanity.com/paper/2605.20563/) (2026)
- [CoAgent: Concurrency Control for Multi-Agent Systems](https://mlsanity.com/paper/2606.15376/) (2026)
- [AWCP: A Workspace Delegation Protocol for Deep-Engagement Collaboration across Remote Agents](https://mlsanity.com/paper/2602.20493/) (2026)
- [ATM: CID-Brokered Pre-Write Admission for Multi-Agent Code Co-Synthesis](https://mlsanity.com/paper/2607.00041/) (2026)
- [Claim Plane: Enforceable Change Intents and Dynamic Scope for Parallel Coding Agents](https://mlsanity.com/paper/2607.21909/) (2026)
- [Formal Architecture Descriptors as Navigation Primitives for AI Coding Agents](https://mlsanity.com/paper/2604.13108/) (2026)
- [AgentRadio: Passive Awareness for Long-Horizon Multi-Agent Collaboration](https://mlsanity.com/paper/2607.28430/) (2026)
- [Claim Plane: Reliability Gains and the Limits of Selective Concurrency for Parallel Coding Agents: A 30-Pair, Three-Seed Confirmatory Study of Deterministic Pre-Write Admission](https://mlsanity.com/paper/2608.00947/) (2026)
- [AgentRoom: Concurrent Multi-Agent Coding in a CRDT-Backed Shared Workspace](https://mlsanity.com/paper/2608.23740/) (2026)
- [When Agent Context Goes Stale: Incoherence in Volatile Agent Context](https://mlsanity.com/paper/2610.05281/) (2026)

## Abstract (arXiv)

Multi-agent LLM systems fail to realize parallel speedups due to costly coordination. We present CodeCRDT, an observation-driven coordination pattern where agents coordinate by monitoring a shared state with observable updates and deterministic convergence, rather than explicit message passing. Using Conflict-Free Replicated Data Types (CRDTs), CodeCRDT enables lock-free, conflict-free concurrent code generation with strong eventual consistency. Evaluation across 600 trials (6 tasks, 50 runs per mode) shows both benefits and trade-offs: up to 21.1% speedup on some tasks, up to 39.4% slowdown on others, and 100% convergence with zero merge failures. The study formalizes observation-driven coordination for stochastic LLM agents, revealing semantic conflict rates (5-10%) and quality-performance tradeoffs, and provides empirical characterization of when parallel coordination succeeds versus fails based on task structure.