# From Query to Counsel: Structured Reasoning with a Multi-Agent Framework and Dataset for Legal Consultation

arXiv: https://arxiv.org/abs/2604.10470 · HTML: https://mlsanity.com/paper/2604.10470/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 15 of its 23 references that are in the map.
- First posted on arXiv 12 Apr 2026.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)
- LLM agents and tool use / [Social simulation and multi-agent debate](https://mlsanity.com/agents/social-simulation-and-multi-agent-debate/)

## Builds on (most cited first, 15 of 15)

- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [BLEURT: Learning Robust Metrics for Text Generation](https://mlsanity.com/paper/2004.04696/) (2020)
- [PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization](https://mlsanity.com/paper/1912.08777/) (2019)
- [How Does NLP Benefit Legal System: A Summary of Legal Artificial Intelligence](https://mlsanity.com/paper/2004.12158/) (2020)
- [Chatlaw: A Multi-Agent Legal Assistant based on a Role-Aligned Mixture-of-Experts Architecture](https://mlsanity.com/paper/2306.16092/) (2023)
- [JEC-QA: A Legal-Domain Question Answering Dataset](https://mlsanity.com/paper/1911.12011/) (2019)
- [LawGPT: A Chinese Legal Knowledge-Enhanced Large Language Model](https://mlsanity.com/paper/2406.04614/) (2024)
- [A Survey of Large Language Model Empowered Agents for Recommendation and Search: Towards Next-Generation Information Retrieval](https://mlsanity.com/paper/2503.05659/) (2025)
- [CaseEncoder: A Knowledge-enhanced Pre-trained Model for Legal Case Encoding](https://mlsanity.com/paper/2305.05393/) (2023)
- [Elevating Legal LLM Responses: Harnessing Trainable Logical Structures and Semantic Knowledge with Legal Reasoning](https://mlsanity.com/paper/2502.07912/) (2025)
- [Legal Question-Answering in the Indian Context: Efficacy, Challenges, and Potential of Modern AI Models](https://mlsanity.com/paper/2309.14735/) (2023)
- [From Newborn to Impact: Bias-Aware Citation Prediction](https://mlsanity.com/paper/2510.19246/) (2025)
- [TPA: Next Token Probability Attribution for Detecting Hallucinations in RAG](https://mlsanity.com/paper/2512.07515/) (2025)
- [Pre-training, Fine-tuning and Re-ranking: A Three-Stage Framework for Legal Question Answering](https://mlsanity.com/paper/2412.19482/) (2024)

## Built on by (most cited first, 3 of 3)

- [Towards Robust Endogenous Reasoning: Unifying Drift Adaptation in Non-Stationary Tuning](https://mlsanity.com/paper/2604.15705/) (2026)
- [PairPref: When Should Memory Guide the Answer? A Benchmark for Contextual Preference Use](https://mlsanity.com/paper/2609.34526/) (2026)
- [Remember Before You're Asked: MemDream for Self-Probing Memory Evolution](https://mlsanity.com/paper/2609.34545/) (2026)

## Abstract (arXiv)

Legal consultation question answering (Legal CQA) presents unique challenges compared to traditional legal QA tasks, including the scarcity of high-quality training data, complex task composition, and strong contextual dependencies. To address these, we construct JurisCQAD, a large-scale dataset of over 43,000 real-world Chinese legal queries annotated with expert-validated positive and negative responses, and design a structured task decomposition that converts each query into a legal element graph integrating entities, events, intents, and legal issues. We further propose JurisMA, a modular multi-agent framework supporting dynamic routing, statutory grounding, and stylistic optimization. Combined with the element graph, the framework enables strong context-aware reasoning, effectively capturing dependencies across legal facts, norms, and procedural logic. Trained on JurisCQAD and evaluated on a refined LawBench, our system significantly outperforms both general-purpose and legal-domain LLMs across multiple lexical and semantic metrics, demonstrating the benefits of interpretable decomposition and modular collaboration in Legal CQA.