FrootAI — AmpliFAI your AI Ecosystem Get Started

FAI Open Source Community

We are building the industry standard for AI primitive unification — together.
Open Source · Free Forever · MIT License · AI primitives and growing.

🌱

Open by Default

Every line of code, every agent, every instruction — open source and MIT licensed. No premium tier, no paywalls, no bait-and-switch.

🤝

Built Together

Primitives created by a passionate community. Your contribution becomes part of the industry standard for AI primitive unification.

🔗

The Binding Glue

FAI Protocol connects what others leave fragmented. MCP handles tools. A2A handles delegation. We handle wiring. One day every AI platform will adopt this.

🏗️

Ship, Don't Talk

We ship production-quality code. Every primitive has WAF alignment, validation, and auto-wiring. No POCs, no demos — real enterprise-grade tools.

Everything is Free — $0 Forever

No premium tier. No enterprise paywall. Every tool, every primitive, every play — open source under MIT license.

  • Solution plays
  • FAI primitives
  • MCP tools
  • VS Code Extension
  • Python SDK + MCP
  • Docker image
  • CLI (npx frootai)
  • Knowledge Modules
  • Comprehensive AI Glossary
  • Cookbook Recipes
  • Agentic Workflows
  • Agent FAI Chatbot

Join the Movement

Ready to Contribute?

Full contribution guide with scaffolding tools, validation checklist, PR templates, and the FAI Factory shipping pipeline. Add agents, skills, instructions, hooks, plugins, workflows, cookbook recipes, or entire solution plays.

Open Contribution Guide →

Community Showcase

Highlighted contributions from the FrootAI community — primitives, plays, integrations, and tutorials.

🧩 Primitives
45

GraphRAG Agent — Knowledge Graph Retrieval

A 238-line agent that combines Neo4j graph traversal with Azure AI Search for multi-hop reasoning. Uses entity extraction to build knowledge graphs from documents, then traverses relationships for context-aware retrieval.

RAGKnowledge GraphNeo4jAgent
by @frootai · Apr 2026
🎯 Solution Plays
120

Play 01: Enterprise RAG — Production Blueprint

The flagship play. 44 files, 4,132 lines. Full DevKit with builder/reviewer/tuner chain, Bicep IaC, eval.py pipeline, 4 instruction files, 4 prompts, and 3 SKILL.md procedures. Deployed to Azure in under 30 minutes.

RAGAzure AI SearchProductionWAF
by @frootai · Apr 2026
🔌 Integrations
35

ServiceNow Integration — ITSM Automation

Full partner integration with 4 MCP tools (ticket_search, ticket_create, ticket_update, cmdb_query), dedicated agent, OAuth2 authentication, and error handling. Compatible with plays 05, 37, 54.

ServiceNowITSMMCP ToolsPartner
by @frootai · Apr 2026
📖 Tutorials
55

From Zero to RAG in 15 Minutes

Step-by-step tutorial: clone the repo, pick Play 01, deploy to Azure with azd up, run evaluation, tune for production. Covers the full builder→reviewer→tuner workflow with real commands.

Getting StartedRAGAzureTutorial
by @frootai · Apr 2026
🏆 Hackathon Winners
65

🏆 May 2026 Winner: Advanced RAG Search Primitives

Hackathon first place. 15 new primitives for advanced RAG patterns: hybrid search, re-ranking, query decomposition, multi-index federation, and citation extraction. All with WAF alignment.

HackathonRAGSearchWinner
by @frootai · May 2026
🧩
Semantic Cache — Redis + Embedding Similarity
Production skill that implements semantic caching for AI responses. Reduces API ...
View
🎯
HIPAA Compliance Kit for Healthcare AI
Complete compliance enrichment for Play 46 (Healthcare Clinical AI). Adds HIPAA/...
View
🎯
Play 22: Multi-Agent Swarm Orchestration
Distributed agent teams with Dapr service mesh. Supervisor-worker topology, conf...
View
🧩
Temporal Workflow Agent — Durable AI Pipelines
Agent for orchestrating long-running AI workflows with Temporal. Saga patterns, ...
View
🧩
Evaluation Pipeline — Azure AI Evaluation SDK
Production skill for building eval.py pipelines. Azure AI Evaluation SDK integra...
View

Built something with FrootAI?

Submit your project to be featured in the community showcase.

Submit to Showcase

The Vision

We are building the industry standard for AI primitive unification. The FAI Protocol is the uniFAIng layer — context-wiring between agents, instructions, skills, hooks, workflows, plugins, tools, prompts, and guardrails. MCP handles tool calling. A2A handles delegation. AG-UI handles rendering. We handle wiring. One day every AI platform will adopt this standard. We are the pioneers.