AgentWiki

Core Concepts

• Planning
• Memory
• Tool Use
• Structured Outputs
• Autonomy
• Agent Safety
• CoALA

Reasoning Techniques

• Chain-of-Thought
• Tree of Thoughts
• Graph of Thoughts
• ReAct
• Reflexion
• Self-Consistency
• Self-Refine
• Chain-of-Verification
• Step-Back
• Least-to-Most
• Skeleton-of-Thought
• Program of Thoughts
• Buffer of Thoughts
• Policy of Thoughts
• LATS
• MCTS + LLM
• Multi-Agent Debate
• Prompt Chaining
• Prompt Engineering
• Latent Reasoning
• Test-Time Compute

Memory Systems

• Hierarchical
• Short-Term
• Long-Term
• CogMem
• Augmentation
• Frameworks
• RAG
• Agentic RAG
• Knowledge Graphs
• Embeddings
• Continual Learning
• Haystack Engineering

Retrieval

• FAISS
• HNSW
• ScaNN

Agent Types

• Autonomous
• ReAct Agents
• Plan & Execute
• Conversational
• Tool-Using
• Code Agents
• Web Browsing
• Deep Search
• Vision Agents
• Voice Agents
• AI Scientists
• GUI Grounding
• Long-Horizon
• Inner Monologue
• Generative Agents
• Multimodal

Design Patterns

• Agent Loop
• Orchestration
• Multi-Agent
• Self-Organizing
• Modular
• Function Calling
• Parallel Calls
• Tool Integration
• Speculative Execution
• Durable Execution
• Context Management
• Mixture of Agents
• Collective Behavior
• Agentic RPA
• Generative UI
• Personalization
• Cost Optimization
• Task Decomposition

Training & Alignment

• RLHF / PPO
• DPO
• Constitutional AI
• RLVR
• Agentic RL
• Process Rewards
• Reasoning Rewards
• Reward Hacking
• Fine-Tuning
• Distillation
• Self-Play
• SPIN
• Self-Evolving
• Trajectory Opt.
• World Models
• PAHF
• Test-Time Adapt.

Frameworks

• MCP
• A2A
• ACP
• ANP
• AG-UI
• LangGraph
• CrewAI
• AutoGen
• DSPy
• LlamaIndex
• OpenAI Agents SDK
• Claude Agent SDK
• Semantic Kernel
• Smolagents
• OpenHands
• MetaGPT
• CAMEL
• TaskWeaver
• Langroid
• Flowise
• LiteLLM
• GraphRAG

Tools & Products

• Devin
• Cursor
• Claude Code
• Aider
• Computer Use
• Bolt.new
• OpenVoiceUI
• Manus AI
• Gorilla
• Voyager

Code & Software

• SWE-Agent
• Semi-Formal Reasoning
• Theory of Code Space
• Repo-Centric Learning
• Agent Context Files

Safety & Security

• Sandbox Security
• Governance
• Threat Modeling
• Identity & Auth
• NIST Standards
• Hallucination
• Tool Attack Chains
• Multi-Turn Jailbreaks
• Proactive Defense

Evaluation

• Overview
• SWE-bench
• WebArena
• AgentBench
• Tau-bench
• Terminal-Bench
• GAIA
• CUB
• Humanity's Last Exam
• World of Workflows
• Instruction Following
• Agent-as-Judge
• LLM-as-Judge
• Agent Index

Research

• Tool Learning Survey
• Formal Verification
• Agentic Skills
• Synthetic Data
• RetroAgent
• Strategy Exploration
• Neurosymbolic
• Small LM Agents
• On-Device Agents

Development

• Agentic Coding
• Debugging
• Observability

Meta

• Wiki API Skill