Part of Build with Claude Using Code and Guides
Engineering Blueprint
Implement Common Agent Architectures
Reference implementations of fundamental agent patterns including prompt chaining, routing, orchestrators, and evaluator-optimizers.
What problem does this solve?
Provides minimal working examples of agent workflows discussed in Anthropic's research, removing guesswork from choosing between architectures. The guide covers when to use prompt chaining (sequential logic), routing (conditional execution), multi-LLM parallelization (diverse tasks), and evaluator-optimizer loops (iterative refinement).
How does it work?
- Review the reference architecture for your use case. 2. Adapt the minimal implementation to your domain. 3. Test the agent loop with sample inputs. 4. Add instrumentation and monitoring as needed. Output: a working agent that solves your problem using proven patterns.
Included Skills
No skills in this group
What's required to run this?
Minimal implementations in Jupyter notebooks. Async multi-agent examples included for parallel coordination.
Key Features
Prompt chaining
Chain multiple LLM calls where output of one feeds into the next
Routing
Use Claude to classify input and route to specialized handlers or prompts
Orchestrator-subagent
Coordinate multiple specialized agents through a master orchestrator
Evaluator-optimizer
Use one LLM to evaluate and improve output from another
About This Blueprint
- Industry
- Technology
More Blueprints to explore
ICP-Filtered Outreach Lists from G2 Buyer Intent
Stop wasting rep time on weak intent signals and mismatched accounts. The workflow ranks ICP-fit buyers by readiness, finds the right contact, and supplies tailored outreach angles for the strongest opportunities.
Check Point Next Generation Firewalls (NGFWs)Ishani D.
Account Manager at G2