Agent Templates

Active

Open Source Claude Code Agent Templates for Enterprise Use Cases

Open source multi-agent workflow templates for Claude Code. Each template implements a real enterprise use case with structured YAML output specs, human review gates, and full audit trails.

Why This Matters

Enterprise AI adoption fails when organisations jump straight to building without a blueprint. These templates provide tested, auditable agent architectures that follow the partial automation model - AI does the work, humans approve it. Download a template, add your knowledge base, and run it.

Focus Areas

Multi-agent orchestrationStructured output via YAML specsHuman-in-the-loop automationEnterprise audit trails

AIAgentsClaude CodeOpen SourceAutomation

Learn more →

Agent Templates is an open source collection of 17 multi-agent workflow templates designed for Claude Code. Each template tackles a specific enterprise use case with a pipeline of specialised sub-agents.

How It Works

Each template follows the same architecture: specialised sub-agents run in sequence, each producing structured YAML output conforming to a spec. A human review gate ensures nothing happens without approval.

Key Design Principles

Partial automation - agents do the work, humans approve it
YAML output specs - every sub-agent produces consistent, validated output against a schema
Auditable - complete pipeline output saved to disk for compliance and cost tracking
Customisable - swap knowledge bases, adjust tone, change models, add policies
Sub-agents stay in their lane - the classifier does not write responses, the researcher does not make policy decisions

Available Templates

Customer Service

Customer Service - Classifier, Researcher, Drafter, and Reviewer sub-agents handle ticket triage through to response drafting
Knowledge Base Q&A - Query Analyser, Retriever, Synthesiser, and Verifier deliver RAG-powered answers with hallucination detection
Sentiment Routing - Emotion Detector, Context Assessor, Priority Scorer, and Router for intelligent ticket escalation

Sales & Marketing

Lead Scoring - Enricher, Scorer, Qualifier, and Router for firmographic enrichment and sales team routing
Content Creation - Brief Analyser, Research Compiler, Content Writer, and Editor for multi-platform marketing content

IT & Engineering

Incident Triage - Classifier, Investigator, Responder, and Reviewer for production incident handling
Code Review - Diff Analyser, Bug Scanner, Style Checker, and Review Composer for AI-powered PR reviews
Doc Generation - Code Analyser, Structure Planner, Doc Writer, and Quality Reviewer for technical documentation

Finance

Invoice Processing - Extractor, Validator, Anomaly Detector, and Approver for invoice and expense automation
Financial Forecasting - Data Ingester, Trend Analyser, Forecaster, and Narrative Writer for variance analysis and projections

HR & People

CV Screening - Parser, Matcher, Assessor, and Ranker with bias mitigation for candidate shortlisting
HR Policy Q&A - Intent Classifier, Policy Retriever, Response Generator, and Compliance Checker for employee self-service

Legal & Compliance

Contract Review - Clause Extractor, Risk Assessor, Comparator, and Summary Writer with risk heat maps
Regulatory Monitoring - Feed Scanner, Change Analyser, Impact Assessor, and Briefing Writer for compliance tracking
Due Diligence - Document Indexer, Risk Extractor, Cross-Referencer, and Report Writer for M&A analysis

Operations & Product

Quality Assurance - Data Parser, Pattern Detector, Root Cause Analyser, and Report Generator with Pareto analysis
PRD Drafting - Input Parser, Research Enricher, Spec Writer, and Reviewer for product requirement documents

Built With

Claude Code CLI with native sub-agent architecture. YAML specs for output consistency. Markdown runbooks and knowledge bases for domain context.