Agent Patterns

Patterns that didn’t break production (with code and trade-offs).

• ReAct Agent

  ★★☆

  Master the ReAct agent pattern in React: reason-act loops, tool use, and guardrails that prevent common failures in production AI workflows.

  Learn the pattern
  → Main article
  Implementations
  • → Python: ReAct Agent implementation

• Task Decomposition Agent

  ★★☆

  Decompose complex goals into manageable subtasks so agents can plan, execute, and verify step by step with better reliability.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Task Decomposition Agent implementation

• Routing Agent

  ★★☆

  Route each request to the best agent or tool using explicit criteria, improving accuracy, speed, and cost efficiency in multi-agent systems.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Routing Agent implementation

• Orchestrator Agent

  ★★☆

  Learn orchestration patterns to delegate subtasks, run executors in parallel, track status, and merge outputs into one reliable final result.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Orchestrator Agent implementation

• Supervisor Agent

  ★★☆

  Use a supervisor agent to validate proposed actions, enforce policies, and block unsafe steps before they impact users or systems.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Supervisor Agent implementation

• Multi-Agent Collaboration

  ★★☆

  Use collaborating agents to split roles, exchange intermediate results, and cross-check outputs to solve complex tasks with higher quality.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Multi-Agent Collaboration implementation

• RAG Agent

  ★★☆

  Build a RAG agent that finds relevant documents, cites sources, and reduces hallucinations in answers.

  Learn the pattern
  → Main article
  Implementations
  • → Python: RAG Agent implementation

• Memory-Augmented Agent

  ★★☆

  Build agents that remember user facts and prior outcomes across sessions to deliver consistent, personalized responses without losing control.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Memory-Augmented Agent implementation

• Reflection Agent

  ★☆☆

  Add one short review pass to catch obvious mistakes before answering, without endless rewrites.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Reflection Agent implementation

• Self-Critique Agent

  ★☆☆

  Run a safe self-critique loop: one schema-based review, one constrained revision, and a change log for stable quality.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Self-Critique Agent implementation

• Fallback Recovery Agent

  ★☆☆

  Build an agent that recovers from tool and model failures with fallback strategies, retries, and controlled degradation.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Fallback Recovery Agent implementation

• Guarded Policy Agent

  ★☆☆

  Implement a policy-gate that allows, denies, rewrites, or escalates risky agent actions for safe and auditable execution.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Guarded Policy Agent implementation

• Code Execution Agent

  ★☆☆

  How an agent runs code in a sandbox to compute reliably, validate hypotheses, and automate tasks with production guardrails.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Code Execution Agent implementation

• Data Analysis Agent

  ★☆☆

  How an agent ingests, cleans, analyzes, and validates data to produce reproducible metrics and conclusions for decisions.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Data Analysis Agent implementation

• Research Agent

  ★☆☆

  Use a bounded research pipeline: search, read, extract facts, and synthesize with citations without tool spam or infinite loops.

  Learn the pattern
  → Main article
  Implementations
  • → Python: Research Agent implementation