Production agents
you can clone today.
Full systems wired end to end. Fork, run, ship — AgentField examples, companion repos, community projects, and tools.
Full systems wired end to end.
Fork, run, rewire. Filter by domain and source — every card links to primitives on Use cases.
Software Development FactorySWE-AF · hundreds of engineering agents that plan, build, test, and ship
A fleet of coordinated AI engineers wired into a topological DAG. Factory control loops, parallel git worktrees, and atomic merges keep the swarm coherent — output lands as PR-grade pull requests, end to end.
AI Security Auditor·SEC-AF · 250 agents that prove vulnerabilities, not flag them
A coordinated agent swarm that traces every vulnerability source-to-sink and adversarially verifies each finding. Confirmed exploits land in your queue — pattern noise never does.
Deep Research Engine·Deep Research · 10,000 agents fan out, evaluate, and recurse on weak branches
Spawns parallel research agents, scores output quality, generates deeper agents on the gaps, and recurses until the answer holds up. Comprehensive, cited, and verifiable.
Document Enrichment Pipeline·Reactive MongoDB · Atlas Triggers wired into a multi-reasoner enrichment pipeline
Documents arrive raw and leave enriched — risk scores, pattern detection, and evidence chains attached. The database becomes the orchestrator; agents react to every write.
Pull Request Reviewer·PR-AF · code review that reads the whole repo from multiple angles
A multi-agent reviewer that indexes the repo, runs parallel review passes from different perspectives — security, performance, style, design — and posts inline comments grounded in real context.
Contract Reviewer·Contract-AF · clause extraction, cross-reference resolution, and adversarial review
Specialized analyst agents extract clauses, resolve cross-references, and an adversary agent stress-tests every claim. Surfaces hidden risk in legal language at production speed.
Clinical Reasoning Suite·Healthcare AF · 11 clinical reasoners with policy gates and PHI boundaries
Healthcare multi-reasoner system spanning 11 clinical intelligence applications. Policy gates on every tool call, PHI never crosses the boundary, and every decision is reproducible end to end.
Cloud Security Scanner·CloudSecurity-AF · shift-left attack path analysis from infrastructure-as-code alone
Reads your Terraform / CloudFormation / Pulumi and reasons about attack paths agents could take through the deployed cloud — before a single resource is provisioned.
Agent Task Planner·PlanDB · local-first SQLite task graphs with dependencies and parallel claiming
A planning database built for agents. Composite task graphs, atomic claiming for parallel execution, mid-flight replanning, and a search-friendly knowledge store — local-first, zero-infra.
Enterprise Simulation Engine·Schema-flexible enterprise simulation with 100+ parallel reasoners
A domain-agnostic simulation system that models any enterprise scenario — pricing, support, marketing, policy — through parallel scenario, actor, behavior, and aggregation layers.
Verified Document Q&A·Hallucination-proof retrieval with ensemble search and parallel claim verification
Production-ready RAG with ensemble retrieval (semantic + keyword + type-specific), confidence-driven iterative refinement, parallel claim verification, and memory-backed caching.
Event-Triggered Agent·One agent, every trigger source — Stripe, GitHub, Slack, cron, HMAC, Bearer webhooks
A reference agent that exercises every built-in trigger source: Stripe payments, GitHub PRs and issues, cron schedules, plus Slack, generic HMAC, and Bearer-auth inbound webhooks.
Vector Memory Starter·TypeScript starter wiring agent-to-agent discovery with vector memory primitives
A TypeScript reasoner demonstrating workflow progress, vector memory set/search, embedding helpers, and agent-to-agent capability discovery — the core coordination primitives in one place.
Fulfillment Triage Agents·Production-grade fulfillment triage
Production-grade multi-agent fulfillment support triage with intake, scoring, routing, and verification agents.
Autonomous IT Operations·Nine-agent IT service management
Nine-agent IT service management system with classification, execution, communication, and learning agents.
Argus Investment System·Investment committee in software
Autonomous financial research system with 5 specialized agents replicating a professional investment committee.
Claims Adjudication·Medical and policy reasoners
Health insurance claims processing with 5 medical and policy reasoners for automated adjudication.
Autonomous Engineering Fleet·SWE-AF, Deep Research, and SEC-AF composed into one engineering system
Composes SWE-AF, Deep Research, and sec-af into a unified autonomous software engineering system.
AF Code Agent·Go SDK code review agent
GitHub code review agent with 12 specialized modules built on the AgentField Go SDK.
Grid Balancing Agent·Energy distribution agents
Automated power grid load balancing with AI-driven energy distribution agents.
AgentCare·Clinical triage and routing
Clinical triage agent for patient risk assessment and care routing.
Codebase QA Agent·Ollama + codebase Q&A
Indexes and answers questions about any codebase using AgentField and Ollama.
Financial Planner·Portfolio analysis agent
Intelligent portfolio analysis and financial planning agent.
Should I Quit·Multi-perspective career advisor
Career decision advisor using specialist swarm agents for multi-perspective analysis.
Music Education Agent·Classical theory tutoring
Classical music education agent with AI-powered lesson planning and theory tutoring.
Kubernetes Deployment·Agents on Kubernetes
Reference deployment of AgentField agents in Kubernetes with full orchestration.
Building something that should live here?
We accept community examples that follow the same boundaries: identity, governance, and verifiable execution. Open a PR and we'll help you ship it.