Skip to content

HKUST-KnowComp/Awesome-Agent-Harness

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
Agent_Harness_Survey.pdf
Agent_Harness_Survey.pdf
 
 
README.md
README.md
 
 
taxonomy.png
taxonomy.png
 
 

Repository files navigation

Awesome Agent Harness 🛠️

Awesome Stars Papers Last Updated License

A curated list of pioneering research papers, tools, and resources on the Agent Harness — the systematic execution layer that transforms raw model capability into sustained, long-horizon autonomy.

arXiv PDF Preview

A Survey on AI Agent Harness

Agent = Model (Stochastic Intelligence) + Harness (Deterministic Infrastructure)

The survey proposes a Unified Architectural Taxonomy that organizes the Agent Harness as a four-layered stack:

  • Layer 1: Execution & Orchestration — The temporal engine driving the autonomous execution loop, model routing, and multi-agent composition.
  • Layer 2: Context & Trajectory Management — The epistemic layer governing state compaction, trajectory persistence, memory hierarchies, and observability.
  • Layer 3: Interaction Surface & Execution Environment — The sensory and actuation organs connecting the agent to the world via tool calling, standardized protocols, and sandboxed execution.
  • Layer 4: Constraints & Guardrails — The independent observer enforcing deterministic laws through access control, permission management, and defense against agent injection.

The figure below illustrates the asymmetric co-evolution between model capability and harness responsibility:

Model-Harness Asymmetric Co-evolution

We aim to provide a comprehensive overview for researchers, developers, and infrastructure engineers interested in this rapidly advancing field.

Contents

Agent Harness Foundations

Cross-layer conceptual works that define and motivate the Agent Harness as a first-class research object.

TitleAuthorYearDescription
Effective harnesses for long-running agents BlogYoung et al.2025long-running agent harness management
Natural-Language Agent HarnessesPan et al.2026natural-language harness design
Harness Engineering for Language Agents: The Harness Layer as Control, Agency, and RuntimeHe et al.2026harness as control, agency, and runtime layer
Building AI Coding Agents for the Terminal: Scaffolding, Harness, Context Engineering, and Lessons LearnedBui et al.2026terminal coding agent scaffolding, context engineering, lessons learned
Harness engineering: leveraging Codex in an agent-first world BlogLopopolo et al.2026Codex-based harness engineering
The importance of Agent Harness in 2026 BlogSchmid et al.2026agent harness importance analysis
What is an agent harness in the context of large-language models? BlogParallel Web Systems et al.2025agent harness concept overview
Meta-Harness: End-to-End Optimization of Model HarnessesLee et al.2026end-to-end automated optimization of harness code

Layer 1: Execution & Orchestration

Acting as the temporal engine of the harness, Layer 1 drives the autonomous execution loop, manages model routing, orchestrates multi-agent compositions, and enforces resilience mechanisms to maintain forward momentum under failures.

Model & Agent Routing

Dynamically determining which LLM or specialized agent should handle a given subtask, optimizing for cost, capability, and resource constraints.

TitleAuthorYearDescription
EvoRoute: Experience-Driven Self-Routing LLM Agent SystemsZhang et al.2026experience-driven self-routing
Best-route: Adaptive llm routing with test-time optimal computeDing et al.2025test-time optimal compute routing
Masrouter: Learning to route llms for multi-agent systemsYue et al.2025multi-agent system routing learning
Adaptive vision-language model routing for computer use agentsLiu et al.2026adaptive VLM routing for computer use
Camar: Continuous actions multi-agent routingPshenitsyn et al.2026continuous-action multi-agent routing
SkillOrchestra: Learning to Route Agents via Skill TransferWang et al.2026skill-transfer-based agent routing
DyTopo: Dynamic Topology Routing for Multi-Agent Reasoning via Semantic MatchingLu et al.2026semantic-matching topology routing
Orchestrating Intelligence: Confidence-Aware Routing for Efficient Multi-Agent Collaboration across Multi-Scale ModelsWang et al.2026confidence-aware multi-scale routing
Learning Query-Aware Budget-Tier Routing for Runtime Agent MemoryZhang et al.2026query-aware budget-tier memory routing
CASTER: Breaking the Cost-Performance Barrier in Multi-Agent Orchestration via Context-Aware Strategy for Task Efficient RoutingLiu et al.2026context-aware task-efficient routing
Budget-aware agentic routing via boundary-guided trainingZhang et al.2026boundary-guided budget-aware routing
ODAR: Principled Adaptive Routing for LLM Reasoning via Active InferenceMa et al.2026active-inference adaptive routing
Optimal-agent-selection: State-aware routing framework for efficient multi-agent collaborationWang et al.2025state-aware optimal agent selection
Towards generalized routing: Model and agent orchestration for adaptive and efficient inferenceGuo et al.2025generalized model-agent orchestration

Multi-Agent Composition & Orchestration

Treating agents as composable, modular entities and orchestrating concurrent subagent spawning, delegation, and synchronized state handoffs.

TitleAuthorYearDescription
Claude Code Subagents BlogAnthropic2025custom AI subagent spawning
Compass: Enhancing agent long-horizon reasoning with evolving contextWan et al.2025evolving context for long-horizon reasoning
Kimi K2. 5: Visual Agentic IntelligenceTeam et al.2026visual agentic intelligence
Swarm: An educational framework exploring ergonomic, lightweight multi-agent orchestration CodeOpenAI et al.2024lightweight multi-agent orchestration
CrewAI: Framework for orchestrating role-playing autonomous AI agents CodeMoura et al.2025role-playing agent orchestration
A Declarative Language for Building And Orchestrating LLM-Powered Agent WorkflowsDaunis et al.2025declarative agent workflow language
Orchestral AI: A Framework for Agent OrchestrationRoman et al.2026general-purpose agent orchestration

Autonomous Loop, Resilience & Human-in-the-Loop

Ensuring the execution loop is resilient to non-termination and drift, and managing the spectrum from full human oversight to closed-loop autonomy.

TitleAuthorYearDescription
Human-in-the-Loop or AI-in-the-Loop? Automate or Collaborate?Natarajan et al.2025human-in-the-loop vs AI-in-the-loop
Adaptive fault tolerance mechanisms of large language models in cloud computing environmentsJin et al.2025adaptive fault tolerance in cloud LLMs
ESAA: Event Sourcing for Autonomous Agents in LLM-Based Software Engineeringdos Santos Filho et al.2026event sourcing for autonomous agents
Combining LLM, Non-monotonic Logical Reasoning, and Human-in-the-loop Feedback in an Assistive AI AgentFu et al.2025LLM + non-monotonic reasoning + HITL
Enabling self-improving agents to learn at test time with human-in-the-loop guidanceHe et al.2025test-time learning with human guidance
Planagent: A multi-modal large language agent for closed-loop vehicle motion planningZheng et al.2026closed-loop vehicle motion planning
A multi-AI agent system for autonomous optimization of agentic AI solutions via iterative refinement and LLM-driven feedback loopsYuksel et al.2025iterative refinement via LLM feedback
Towards LLM-enabled autonomous combustion research: A literature-aware agent for self-corrective modeling workflowsXiao et al.2026autonomous combustion research agent
From llm reasoning to autonomous ai agents: A comprehensive reviewFerrag et al.2025LLM reasoning to autonomous agents survey

Layer 2: Context & Trajectory Management

While the orchestration layer manages execution time, Layer 2 governs the agent's epistemic space — mitigating context window saturation, catastrophic forgetting, and maintaining strict observability.

Memory Systems

Structured, queryable knowledge layers ranging from production-ready platforms to research prototypes.

TitleAuthorYearDescription
Mem0: Building production-ready ai agents with scalable long-term memoryChhikara et al.2025scalable production long-term memory
Zep: a temporal knowledge graph architecture for agent memoryRasmussen et al.2025temporal knowledge graph memory
Memory is all you need: Testing how model memory affects llm performance in annotation tasksTimoneda et al.2025memory effects on LLM annotation
AMemGym: Interactive Memory Benchmarking for Assistants in Long-Horizon ConversationsJiayang et al.2026interactive memory benchmarking
Evaluating memory in llm agents via incremental multi-turn interactionsHu et al.2025incremental multi-turn memory eval
Nemori: Self-organizing agent memory inspired by cognitive scienceNan et al.2025self-organizing cognitive memory
MemGPT: towards LLMs as operating systems.Packer et al.2023LLM as operating system with memory tiers
A-mem: Agentic memory for llm agentsXu et al.2025agentic self-organizing memory
Memagent: Reshaping long-context llm with multi-conv rl-based memory agentYu et al.2025multi-conv RL-based memory agent
G-memory: Tracing hierarchical memory for multi-agent systemsZhang et al.2025hierarchical multi-agent memory tracing
Hipporag: Neurobiologically inspired long-term memory for large language modelsGutierrez et al.2024neurobiological long-term memory
SimpleMem: Efficient Lifelong Memory for LLM AgentsLiu et al.2026efficient lifelong memory for agents
General agentic memory via deep researchYan et al.2025agentic memory via deep research
Choosing How to Remember: Adaptive Memory Structures for LLM AgentsLu et al.2026adaptive memory structure selection
From Lossy to Verified: A Provenance-Aware Tiered Memory for AgentsZhu et al.2026provenance-aware tiered memory
Lifelong learning of large language model based agents: A roadmapZheng et al.2026lifelong learning roadmap for agents
Memory Poisoning Attack and Defense on Memory Based LLM-AgentsSunil et al.2026memory poisoning attack and defense

Context Compression

Strategies to prevent Context Rot — the progressive degradation of reasoning quality due to accumulated irrelevant tokens.

TitleAuthorYearDescription
Acon: Optimizing context compression for long-horizon llm agentsKang et al.2025long-horizon agent context compression
Longllmlingua: Accelerating and enhancing llms in long context scenarios via prompt compressionJiang et al.2024prompt compression via token scoring
Scaling llm multi-turn rl with end-to-end summarization-based context managementLu et al.2025summarization-based context management
Longcodebench: Evaluating coding llms at 1m context windowsRando et al.20251M-token coding evaluation
Scaling long-horizon llm agent via context-foldingSun et al.2025hierarchical trajectory folding
Pretraining context compressor for large language models with embedding-based memoryDai et al.2025embedding-based context compressor
SWE-Pruner: Self-Adaptive Context Pruning for Coding AgentsWang et al.2026self-adaptive context pruning for code
The Complexity Trap: Simple Observation Masking Is as Efficient as LLM Summarization for Agent Context ManagementLindenbauer et al.2025observation masking vs summarization
Longcodezip: Compress long context for code language modelsShi et al.2025long-context code compression
Mem1: Learning to synergize memory and reasoning for efficient long-horizon agentsZhou et al.2025synergize memory and reasoning
ContextBench: A Benchmark for Context Retrieval in Coding AgentsLi et al.2026context retrieval benchmarking
Memrl: Self-evolving agents via runtime reinforcement learning on episodic memoryZhang et al.2026runtime RL on episodic memory
Hiagent: Hierarchical working memory management for solving long-horizon agent tasks with large language modelHu et al.2025hierarchical working memory management
SWE Context Bench: A Benchmark for Context Learning in CodingZhu et al.2026context learning benchmarking
Safesieve: From heuristics to experience in progressive pruning for llm-based multi-agent communicationZhang et al.2026progressive multi-agent comm pruning

Trajectory Persistence & Observability

Persisting the agent's execution history to external storage for recovery, replay, and continuous learning, while decoupling observability from the model's working memory.

TitleAuthorYearDescription
Reducing Cost of LLM Agents with Trajectory ReductionXiao et al.2025trajectory reduction for efficiency
Semantic Checkpointing for Stateless LLM Agents in Multi-Tenant Enterprise SystemsRoshan et al.2025semantic checkpointing for stateless agents
Large-scale Evaluation of Notebook Checkpointing with AI AgentsFang et al.2025notebook checkpointing evaluation
AgentTrace: A Structured Logging Framework for Agent System ObservabilityAlSayyad et al.2026structured logging for observability
AgentSight: System-Level Observability for AI Agents Using eBPFZheng et al.2025eBPF-based system-level observability
Durable Execution in LangGraph BlogLangChain et al.2026fault-tolerant durable execution

Self-Evolving Architectures

Agent systems that improve their own capabilities, prompts, or memory structures at test time or through continuous interaction.

TitleAuthorYearDescription
Darwin godel machine: Open-ended evolution of self-improving agentsZhang et al.2025open-ended self-improving evolution
Your agent may misevolve: Emergent risks in self-evolving llm agentsShao et al.2025emergent risks in self-evolution
Live-SWE-agent: Can Software Engineering Agents Self-Evolve on the Fly?Xia et al.2025on-the-fly SWE agent self-evolution
Agentic context engineering: Evolving contexts for self-improving language modelsZhang et al.2025evolving contexts for self-improvement
Gepa: Reflective prompt evolution can outperform reinforcement learningAgrawal et al.2025reflective prompt evolution
Dynamic cheatsheet: Test-time learning with adaptive memorySuzgun et al.2026test-time learning with adaptive memory
Adaptive self-improvement llm agentic system for ml library developmentZhang et al.2025self-improvement for ML library dev
AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel OptimizationZhang et al.2025self-improving kernel optimization
Memento: Fine-tuning LLM Agents without Fine-tuning LLMsZhou et al.2025fine-tuning agents without LLM FT
Multi-agent evolve: Llm self-improve through co-evolutionChen et al.2025LLM self-improve through co-evolution
Ragen: Understanding self-evolution in llm agents via multi-turn reinforcement learningWang et al.2025self-evolution via multi-turn RL
Evo-memory: Benchmarking llm agent test-time learning with self-evolving memoryWei et al.2025benchmarking test-time self-evolving memory
WebEvolver: Enhancing Web Agent Self-Improvement with Co-evolving World ModelFang et al.2025co-evolving web world model
SEAgent: Self-Evolving Computer Use Agent with Autonomous Learning from ExperienceSun et al.2025self-evolving computer use agent
Self-evolving multi-agent simulations for realistic clinical interactionsAlmansoori et al.2025self-evolving clinical simulations
EvoAgent: Self-evolving Agent with Continual World Model for Long-Horizon TasksFeng et al.2025continual world model for long-horizon
Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero DataAcikgoz et al.2026self-evolving tool learning from zero
EvoTool: Self-evolving tool-use policy optimization in llm agents via blame-aware mutation and diversity-aware selectionYang et al.2026blame-aware tool-use optimization
AutoSkill: Experience-Driven Lifelong Learning via Skill Self-EvolutionYang et al.2026experience-driven lifelong skill evolution
MemSkill: Learning and Evolving Memory Skills for Self-Evolving AgentsZhang et al.2026learning and evolving memory skills
EvoConfig: Self-Evolving Multi-Agent Systems for Efficient Autonomous Environment ConfigurationGuo et al.2026self-evolving multi-agent configuration
Over-Searching in Search-Augmented Large Language ModelsXie et al.2026over-searching in search-augmented LLMs

Agentic Skills

Modular, reusable capabilities that agents acquire, compose, and execute to extend their action space.

TitleAuthorYearDescription
Inducing programmatic skills for agentic tasksWang et al.2025inducing programmatic skills
SkillCraft: Can LLM Agents Learn to Use Tools Skillfully?Chen et al.2026tool-use skill learning evaluation
SoK: Agentic Skills--Beyond Tool Use in LLM AgentsJiang et al.2026systematization of agentic skills
SkillReducer: Optimizing LLM Agent Skills for Token EfficiencyGao et al.2026token-efficient skill optimization
Agent skills for large language models: Architecture, acquisition, security, and the path forwardXu et al.2026skill architecture, acquisition, security
Agent Skills: A Data-Driven Analysis of Claude Skills for Extending Large Language Model FunctionalityLing et al.2026data-driven Claude skill analysis
SkillRouter: Retrieve-and-Rerank Skill Selection for LLM Agents at ScaleZheng et al.2026retrieve-and-rerank skill selection
When single-agent with skills replace multi-agent systems and when they failLi et al.2026single-agent skills vs multi-agent
EvoSkill: Automated Skill Discovery for Multi-Agent SystemsAlzubi et al.2026automated multi-agent skill discovery
SkillsBench: Benchmarking how well agent skills work across diverse tasksLi et al.2026skill benchmarking across diverse tasks
Cua-skill: Develop skills for computer using agentChen et al.2026skills for computer-using agents
Introducing Agent Skills BlogAnthropic et al.2025agent skill platform launch
Reinforcement Learning for Self-Improving Agent with Skill LibraryWang et al.2025RL-based self-improving skill library
Agentic Proposing: Enhancing Large Language Model Reasoning via Compositional Skill SynthesisJiao et al.2026compositional skill synthesis
SkillClaw: Let Skills Evolve Collectively with Agentic EvolverMa et al.2026collective skill evolution via cloud sharing

Skills Security

Security vulnerabilities and defenses related to agentic skill systems and skill-based prompt injection.

TitleAuthorYearDescription
Agent Skills Enable a New Class of Realistic and Trivially Simple Prompt InjectionsSchmotz et al.2025skill-based prompt injection analysis
Agent Skills in the Wild: An Empirical Study of Security Vulnerabilities at ScaleLiu et al.2026skill security vulnerabilities at scale
Malicious Agent Skills in the Wild: A Large-Scale Security Empirical StudyLiu et al.2026malicious skill detection study
When Skills Lie: Hidden-Comment Injection in LLM AgentsWang et al.2026hidden-comment skill injection
Zombie Agents: Persistent Control of Self-Evolving LLM Agents via Self-Reinforcing InjectionsYang et al.2026persistent control via self-reinforcing injection

Layer 3: Interaction Surface & Execution Environment

Because language models are inherently disembodied, Layer 3 constitutes the sensory and actuation organs of the agentic system — standardizing interfaces for tool calling and code execution, and enforcing hardware-level isolation.

Standardized Protocols & Interaction Surface

Defining and standardizing how agents interact with tools, APIs, and external environments.

TitleAuthorYearDescription
From language to action: a review of large language models as autonomous agents and tool usersChowa et al.2026LLM as autonomous agent review
Defining and Detecting the Defects of Large Language Model-based Autonomous AgentsNing et al.2026LLM agent defect detection
Llm agents making agent toolsWolflein et al.2025agents making agent tools
Code-Mode: Plug-and-play library to enable agents to call MCP and UTCP tools via code execution CodeProtocol et al.2026MCP/UTCP via code execution
Ui-tars: Pioneering automated gui interaction with native agentsQin et al.2025native automated GUI interaction
GeoJSON agents: a multi-agent LLM architecture for geospatial analysis—function calling vs. code generationLuo et al.2026function calling vs code generation
Beyond Perfect APIs: A Comprehensive Evaluation of LLM Agents Under Real-World API ComplexityKim et al.2026real-world API complexity evaluation
Beyond Message Passing: Toward Semantically Aligned Agent CommunicationYuan et al.2026semantically aligned agent communication
Improving Google A2A Protocol: Protecting Sensitive Data and Mitigating Unintended Harms in Multi-Agent SystemsLouck et al.2025A2A protocol sensitive data protection
A2ASecBench: A Protocol-Aware Security Benchmark for Agent-to-Agent Multi-Agent SystemsLi et al.2026protocol-aware A2A security benchmark

Tool Use & Code Execution

Benchmarks and methods for evaluating and improving agent tool use capabilities.

TitleAuthorYearDescription
WebOperator: Action-Aware Tree Search for Autonomous Agents in Web EnvironmentDihan et al.2025action-aware web tree search
Terminal-bench: Benchmarking agents on hard, realistic tasks in command line interfacesMerrill et al.2026CLI task benchmarking
Budget-Constrained Agentic Large Language Models: Intention-Based Planning for Costly Tool UseLiu et al.2026budget-constrained tool planning
Toolsandbox: A stateful, conversational, interactive evaluation benchmark for llm tool use capabilitiesLu et al.2025stateful tool-use evaluation

Layer 4: Constraints & Guardrails

Because LLM outputs are inherently probabilistic, Layer 4 acts as an independent observer and judge — imposing deterministic laws of physics and security boundaries on the system, operating entirely out-of-band.

Sandboxing & Execution Environments

Isolating agent execution to contain erratic behaviors and protect host infrastructure.

TitleAuthorYearDescription
The two patterns by which agents connect sandboxes BlogLangChain et al.2026agent-sandbox connection patterns
SWE-MiniSandbox: Container-Free Reinforcement Learning for Building Software Engineering AgentsYuan et al.2026container-free RL sandbox
Ui-tars-2 technical report: Advancing gui agent with multi-turn reinforcement learningWang et al.2025multi-turn RL GUI agent sandbox
Computer Environments Elicit General Agentic Intelligence in LLMsCheng et al.2026sandbox for agentic intelligence
Sari Sandbox: A Virtual Retail Store Environment for Embodied AI AgentsGajo et al.2025virtual retail store environment
SandboxSocial: A Sandbox for Social Media Using Multimodal AI AgentsTouzel et al.2025social media simulation sandbox
cellmate: Sandboxing browser ai agentsMeng et al.2025browser AI agent sandboxing
AgentBay: A Hybrid Interaction Sandbox for Seamless Human-AI Intervention in Agentic SystemsPiao et al.2025hybrid human-AI interaction sandbox
Static Sandboxes Are Inadequate: Modeling Societal Complexity Requires Open-Ended Co-Evolution in LLM-Based Multi-Agent SimulationsChen et al.2025static sandboxes are inadequate; open-ended co-evolution needed
Deepresearchgym: A free, transparent, and reproducible evaluation sandbox for deep researchCoelho et al.2025reproducible deep research evaluation
SWE-World: Building Software Engineering Agents in Docker-Free EnvironmentsSun et al.2026Docker-free SWE environments
R2e-gym: Procedural environments and hybrid verifiers for scaling open-weights swe agentsJain et al.2025procedural environments with hybrid verifiers

Governance Boundaries

Enforcing access control, permission management, and policy compliance for agent actions.

TitleAuthorYearDescription
POLARIS: Typed Planning and Governed Execution for Agentic AI in Back-Office AutomationMoslemi et al.2026typed planning and governed execution
ContextCov: Deriving and Enforcing Executable Constraints from Agent Instruction FilesSharma et al.2026executable constraint enforcement
Sandbox-runtime: A lightweight sandboxing tool for enforcing filesystem and network restrictions on arbitrary processes at the OS level, without requiring a container CodeAnthropic et al.2026OS-level filesystem/network sandboxing
Securing AI Agent ExecutionBuhler et al.2025agent execution security analysis
BashArena: A Control Setting for Highly Privileged AI AgentsKaufman et al.2025highly-privileged agent control setting
Breaking the Protocol: Security Analysis of the Model Context Protocol Specification and Prompt Injection Vulnerabilities in Tool-Integrated LLM AgentsMaloyan et al.2026MCP specification security analysis

Agent Injection & Defense

Defending against adversarial prompt injection attacks targeting agentic systems.

TitleAuthorYearDescription
Skill-Inject: Measuring Agent Vulnerability to Skill File AttacksSchmotz et al.2026skill file attack measurement
AgentDyn: A Dynamic Open-Ended Benchmark for Evaluating Prompt Injection Attacks of Real-World Agent Security SystemLi et al.2026dynamic prompt injection benchmark
WebSentinel: Detecting and Localizing Prompt Injection Attacks for Web AgentsWang et al.2026web agent injection detection
ReasAlign: Reasoning Enhanced Safety Alignment against Prompt Injection AttackLi et al.2026reasoning-enhanced injection safety
From assistant to double agent: Formalizing and benchmarking attacks on openclaw for personalized local ai agentWang et al.2026formalizing attacks on OpenClaw for personalized local agents
Taming OpenClaw: Security Analysis and Mitigation of Autonomous LLM Agent ThreatsDeng et al.2026OpenClaw security analysis and mitigation
Uncovering Security Threats and Architecting Defenses in Autonomous Agents: A Case Study of OpenClawYing et al.2026OpenClaw threat architecture and defenses
ClawKeeper: Comprehensive Safety Protection for OpenClaw Agents Through Skills, Plugins, and WatchersLiu et al.2026comprehensive safety via skills, plugins, and watchers
A trajectory-based safety audit of clawdbot (openclaw)Chen et al.2026trajectory-based safety audit
OpenClaw PRISM: A Zero-Fork, Defense-in-Depth Runtime Security Layer for Tool-Augmented LLM AgentsLi et al.2026zero-fork defense-in-depth runtime
OpenClaw Agents on Moltbook: Risky Instruction Sharing and Norm Enforcement in an Agent-Only Social NetworkManik et al.2026risky instruction sharing and norm enforcement in agent networks

Contributing

Contributions are welcome! To add a paper, open a pull request with the new entry added to the relevant section, following the format below:

Title-Year-Brief description

Please ensure the paper is directly relevant to the Agent Harness infrastructure.

This repository is maintained in conjunction with the survey paper "A Survey on AI Agent Harness".

About

No description, website, or topics provided.

Resources

Readme
Activity
Custom properties

Stars

61 stars

Watchers

3 watching

Forks

6 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors