A survey that unifies prior work on multi-agent LLM systems via the LIFE framework, mapping dependencies across collaboration, failure attribution, and autonomous self-evolution while identifying cross-stage challenges.
hub Canonical reference
Plan-and-Act: Improving Planning of Agents for Long-Horizon Tasks
Canonical reference. 92% of citing Pith papers cite this work as background.
28
Pith papers citing it
1
external citations · Pith
Background
92% of classified citations
abstract
Large language models (LLMs) have shown remarkable advancements in enabling language agents to tackle simple tasks. However, applying them for complex, multi-step, long-horizon tasks remains a challenge. Recent work have found success by separating high-level planning from low-level execution, which enables the model to effectively balance high-level planning objectives and low-level execution details. However, generating accurate plans remains difficult since LLMs are not inherently trained for this task. To address this, we propose Plan-and-Act, a novel framework that incorporates explicit planning into LLM-based agents and introduces a scalable method to enhance plan generation through a novel synthetic data generation method. Plan-and-Act consists of a Planner model which generates structured, high-level plans to achieve user goals, and an Executor model that translates these plans into environment-specific actions. To train the Planner effectively, we introduce a synthetic data generation method that annotates ground-truth trajectories with feasible plans, augmented with diverse and extensive examples to enhance generalization. We evaluate Plan-and-Act using web navigation as a representative long-horizon planning environment, demonstrating a state-of-the-art 57.58% success rate on the WebArena-Lite benchmark as well as a text-only state-of-the-art 81.36% success rate on WebVoyager.
hub tools
citation-role summary
background 12
baseline 1
citation-polarity summary
representative citing papers
SDP constructs a task-induced state space from raw text by having agents commit to and certify natural-language predicates as states, enabling structured planning and analysis in unstructured language environments.
ActGuide-RL uses human action data as plan-style guidance in mixed-policy RL to overcome exploration barriers in LLM agents, matching SFT+RL performance on search benchmarks without cold-start training.
FlowSteer is a prompt-only attack that biases multi-agent LLM workflow planning to propagate malicious signals, raising success rates by up to 55%, with FlowGuard as an input-side defense reducing it by up to 34%.
OMC framework turns multi-agent AI into self-organizing companies with Talents, Talent Market, and E²R search, achieving 84.67% success on PRDBench (15.48 points above prior art).
Long-horizon LLM agents exhibit a finite-grid phase transition in world-model fidelity under state cardinality and dependency density, with world-state collapse preceding action failure.
Mastermind's dual-loop planner learns transferable strategies via SFT and milestone GRPO, raising GPT-5.5 executor pass rate on 200 held-out CyberGym tasks from 60% to 84.5%.
MetaPS trains models via simulation rollouts to select from programmatic strategy libraries for market agents, yielding better performance than fixed or direct LLM baselines across model sizes.
The paper defines a five-plane reference architecture for runtime governance of production AI agents that enforces policies on delegated actions via reasoning and enforcement planes, six interruption primitives, four correctness invariants, and a reference implementation showing microsecond adjudica
StepFinder turns execution logs into temporal semantic sequences via LLMs then uses temporal modeling plus attention to attribute failures to specific steps more accurately and 79% faster than direct LLM methods on the Who&When benchmark.
A single long-horizon agent with dependency-aware traversal and RepoMemory produces more complete, consistent, hierarchical repository documentation than open and closed baselines.
No existing AI security framework covers a majority of the 193 identified multi-agent system threats in any category, with OWASP Agentic Security Initiative achieving the highest overall coverage at 65.3%.
HiMAC decomposes LLM agent tasks into macro planning and micro execution using critic-free hierarchical RL and iterative co-evolution, outperforming baselines on ALFWorld, WebShop, and Sokoban.
KGLAMP uses a dynamically updated knowledge graph to guide LLMs in creating and replanning PDDL specifications for heterogeneous multi-robot teams, reporting at least 25.3% better performance than LLM-only or classical PDDL baselines on the MAT-THOR benchmark.
Evo-Memory is a new streaming benchmark and evaluation framework for self-evolving memory in LLM agents, unifying over ten memory modules and introducing the ReMem pipeline for continual improvement on multi-turn and reasoning datasets.
AlphaCast is a training-free LLM framework that performs interactive multi-stage reasoning for time series forecasting by integrating feature extraction, knowledge bases, case libraries, and contextual pools.
TabClaw is an interactive LLM agent for spreadsheets that exposes editable plans, uses parallel specialist agents, streams ReAct loops, and distills skills from user feedback, reporting improved benchmark task completion.
AMREC couples molecule-aware mismatch tracking with expanded candidate exploration and trajectory-level selection to achieve stronger recovery of molecular identity from invalid ChEBI-20 LLM drafts than prior repair strategies.
A survey that organizes existing work on LLM-based agents around code as the central harness, structured in three layers of interfaces, mechanisms, and multi-agent scaling, with applications across domains and listed open challenges.
StraTA improves LLM agent success rates to 93.1% on ALFWorld and 84.2% on WebShop by sampling a compact initial strategy and training it jointly with action execution via hierarchical GRPO-style rollouts.
The paper delivers the first comprehensive overview of RL for GUI agents, organizing methods into offline, online, and hybrid strategies while analyzing trends in rewards, efficiency, and deliberation to outline a future roadmap.
AdaPlan-H enables LLM agents to generate self-adaptive hierarchical plans that adjust detail level to task difficulty, improving success rates in multi-step tasks.
An end-to-end LLM framework refines natural language into valid PDDL domains and problems via hardcoded and dynamic agents, generates plans with standard engines, and returns readable output.
Introduces a migration framework and A/B/C taxonomy to convert static expert LLM workflows into self-evolution-ready harnesses.
citing papers explorer
-
Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systems
A survey that unifies prior work on multi-agent LLM systems via the LIFE framework, mapping dependencies across collaboration, failure attribution, and autonomous self-evolution while identifying cross-stage challenges.
-
State-Centric Decision Process
SDP constructs a task-induced state space from raw text by having agents commit to and certify natural-language predicates as states, enabling structured planning and analysis in unstructured language environments.
-
Learning Agentic Policy from Action Guidance
ActGuide-RL uses human action data as plan-style guidance in mixed-policy RL to overcome exploration barriers in LLM agents, matching SFT+RL performance on search benchmarks without cold-start training.
-
FlowSteer: Prompt-Only Workflow Steering Exposes Planning-Time Vulnerabilities in Multi-Agent LLM Systems
FlowSteer is a prompt-only attack that biases multi-agent LLM workflow planning to propagate malicious signals, raising success rates by up to 55%, with FlowGuard as an input-side defense reducing it by up to 34%.
-
From Skills to Talent: Organising Heterogeneous Agents as a Real-World Company
OMC framework turns multi-agent AI into self-organizing companies with Talents, Talent Market, and E²R search, achieving 84.67% success on PRDBench (15.48 points above prior art).
-
World-Model Collapse as a Phase Transition
Long-horizon LLM agents exhibit a finite-grid phase transition in world-model fidelity under state cardinality and dependency density, with world-state collapse preceding action failure.
-
Mastermind: Strategy-grounded Learning for Repository-Scale Vulnerability Reproduction
Mastermind's dual-loop planner learns transferable strategies via SFT and milestone GRPO, raising GPT-5.5 executor pass rate on 200 held-out CyberGym tasks from 60% to 84.5%.
-
MetaPS: Adaptive Programmatic Strategy Selection for Market Agents
MetaPS trains models via simulation rollouts to select from programmatic strategy libraries for market agents, yielding better performance than fixed or direct LLM baselines across model sizes.
-
A Five-Plane Reference Architecture for Runtime Governance of Production AI Agents
The paper defines a five-plane reference architecture for runtime governance of production AI agents that enforces policies on delegated actions via reasoning and enforcement planes, six interruption primitives, four correctness invariants, and a reference implementation showing microsecond adjudica
-
StepFinder: A Temporal Semantic Framework for Failure Attribution in Multi-Agent Systems
StepFinder turns execution logs into temporal semantic sequences via LLMs then uses temporal modeling plus attention to attribute failures to specific steps more accurately and 79% faster than direct LLM methods on the Who&When benchmark.
-
Remember Your Trace: Memory-Guided Long-Horizon Agentic Framework for Consistent and Hierarchical Repository-Level Code Documentation
A single long-horizon agent with dependency-aware traversal and RepoMemory produces more complete, consistent, hierarchical repository documentation than open and closed baselines.
-
Security Considerations for Multi-agent Systems
No existing AI security framework covers a majority of the 193 identified multi-agent system threats in any category, with OWASP Agentic Security Initiative achieving the highest overall coverage at 65.3%.
-
HiMAC: Hierarchical Macro-Micro Learning for Long-Horizon LLM Agents
HiMAC decomposes LLM agent tasks into macro planning and micro execution using critic-free hierarchical RL and iterative co-evolution, outperforming baselines on ALFWorld, WebShop, and Sokoban.
-
KGLAMP: Knowledge Graph-guided Language model for Adaptive Multi-robot Planning and Replanning
KGLAMP uses a dynamically updated knowledge graph to guide LLMs in creating and replanning PDDL specifications for heterogeneous multi-robot teams, reporting at least 25.3% better performance than LLM-only or classical PDDL baselines on the MAT-THOR benchmark.
-
Evo-Memory: Benchmarking LLM Agent Test-time Learning with Self-Evolving Memory
Evo-Memory is a new streaming benchmark and evaluation framework for self-evolving memory in LLM agents, unifying over ten memory modules and introducing the ReMem pipeline for continual improvement on multi-turn and reasoning datasets.
-
AlphaCast: A Human Wisdom-LLM Intelligence Co-Reasoning Framework for Interactive Time Series Forecasting
AlphaCast is a training-free LLM framework that performs interactive multi-stage reasoning for time series forecasting by integrating feature extraction, knowledge bases, case libraries, and contextual pools.
-
TabClaw: An Interactive and Self-Evolving Agent for Spreadsheet Manipulation and Table Reasoning
TabClaw is an interactive LLM agent for spreadsheets that exposes editable plans, uses parallel specialist agents, streams ReAct loops, and distills skills from user feedback, reporting improved benchmark task completion.
-
Agentic Molecular Recovery via Molecule-Aware Exploration
AMREC couples molecule-aware mismatch tracking with expanded candidate exploration and trajectory-level selection to achieve stronger recovery of molecular identity from invalid ChEBI-20 LLM drafts than prior repair strategies.
-
Code as Agent Harness
A survey that organizes existing work on LLM-based agents around code as the central harness, structured in three layers of interfaces, mechanisms, and multi-agent scaling, with applications across domains and listed open challenges.
-
StraTA: Incentivizing Agentic Reinforcement Learning with Strategic Trajectory Abstraction
StraTA improves LLM agent success rates to 93.1% on ALFWorld and 84.2% on WebShop by sampling a compact initial strategy and training it jointly with action execution via hierarchical GRPO-style rollouts.
-
GUI Agents with Reinforcement Learning: Toward Digital Inhabitants
The paper delivers the first comprehensive overview of RL for GUI agents, organizing methods into offline, online, and hybrid strategies while analyzing trends in rewards, efficiency, and deliberation to outline a future roadmap.
-
From Coarse to Fine: Self-Adaptive Hierarchical Planning for LLM Agents
AdaPlan-H enables LLM agents to generate self-adaptive hierarchical plans that adjust detail level to task difficulty, improving success rates in multi-step tasks.
-
End-to-end PDDL Planning with Hardcoded and Dynamic Agents
An end-to-end LLM framework refines natural language into valid PDDL domains and problems via hardcoded and dynamic agents, generates plans with standard engines, and returns readable output.
-
Toward Self-Evolution-Ready Workflow Harnesses: A Reversible Migration Path and Convertibility Taxonomy for Expert LLM Pipelines
Introduces a migration framework and A/B/C taxonomy to convert static expert LLM workflows into self-evolution-ready harnesses.
-
Building Customer Support AI Agents at 100M-User Scale: An Evaluation-Driven Framework
An evaluation-driven framework for customer support AI agents at Nubank integrates context engineering, LLM judges, and A/B testing to deliver up to 37pp NPS gains and strong offline-online correlation across five production domains.
-
MimirRAG: A Multi-Agent RAG Framework for Financial Data Retrieval with Metadata Integration
MimirRAG, a multi-agent RAG framework with metadata integration and table-aware chunking, reaches 89.3% accuracy on FinanceBench and outperforms prior baselines for financial document retrieval.
-
Agentic Reasoning for Large Language Models
The survey structures agentic reasoning for LLMs into foundational, self-evolving, and collective multi-agent layers while distinguishing in-context orchestration from post-training optimization and reviewing applications across domains.
- ReCodeAgent: A Multi-agent Workflow for Language-Agnostic Translation and Validation of Large-Scale Repositories