Aixgo Features

Complete Technical Documentation:
FEATURES.md on GitHub - Authoritative feature catalog with code references
PATTERNS.md on GitHub - Deep-dive guides for all 13 orchestration patterns
Roadmap - Development roadmap and planned features

At a glance

Category	Count	Status
Agent Types	6 specialized types	Implemented
LLM Providers	8 cloud + local	Implemented
Orchestration Patterns	13 patterns	Implemented
Security Modes	4 auth modes	Implemented
Observability Backends	6+ backends	Implemented

Performance

Binary Size: <20MB
Cold Start: <100ms
Infrastructure Savings: 60-70% vs Python frameworks

Agent types

Build specialized agents for different tasks:

Agent	Purpose	Key Features
ReAct	Reasoning + Acting	Tool calling, streaming, structured outputs
Classifier	Content routing	Confidence scores, multi-label, custom taxonomies
Aggregator	Multi-agent synthesis	5 LLM strategies + 4 voting modes
Planner	Task decomposition	Dependency analysis, progress tracking
Producer	Message generation	Interval-based, event-driven
Logger	Audit trails	Structured JSON, multiple targets

LLM providers

Connect to any provider with a unified interface:

Provider	Models	Status
OpenAI	GPT-4, GPT-3.5 Turbo	Implemented
Anthropic	Claude 3.5 Sonnet, Opus, Haiku	Implemented
Google Gemini	Gemini 1.5 Pro, Flash	Implemented
xAI	Grok-beta	Implemented
Vertex AI	Gemini on GCP	Implemented
HuggingFace	Meta-Llama, Mistral, 100+ models	Implemented
Ollama	phi, llama, mistral, gemma (local)	Implemented
vLLM	Self-hosted inference	Implemented

Orchestration patterns

All 13 patterns are production-ready:

Pattern	Benefit
Supervisor	Simple multi-agent coordination
Sequential	ETL and content pipelines
Parallel	3-4× speedup
Router	25-50% cost savings
Swarm	Adaptive agent handoffs
Hierarchical	Complex workflows
RAG	70% token reduction
Reflection	20-50% quality improvement
Ensemble	25-50% error reduction
Classifier	Intent-based routing
Aggregation	Expert consensus
Planning	Multi-step workflows
MapReduce	Large dataset processing

View Pattern Details →

Security

Enterprise-grade security built-in:

4 Auth Modes: Disabled, Delegated (IAP), Builtin (API keys), Hybrid
Input Protection: SSRF protection, sanitization, prompt injection defense
Rate Limiting: Token bucket, per-user quotas
Audit: SIEM integration (Elasticsearch, Splunk, Datadog)

Observability

Complete production monitoring:

Tracing: OpenTelemetry with Langfuse, Jaeger, Honeycomb, Grafana
Metrics: Prometheus export, system and agent metrics
Cost Tracking: Automatic token counting, per-request costs
Health Checks: Kubernetes-ready liveness and readiness probes

Deployment

Deploy anywhere with Go’s simplicity:

Target	Details
Single Binary	<20MB, zero dependencies
Docker	Multi-stage builds, ~50MB standard
Kubernetes	Full manifests, HPA ready
Cloud Run	Auto-scaling, IAP integration

Getting started

bash

go get github.com/aixgo-dev/aixgo

yaml

# config/agents.yaml
supervisor:
  name: coordinator
  model: gpt-4-turbo

agents:
  - name: analyzer
    role: react
    model: gpt-4-turbo
    prompt: "You are a data analyst."

package main

import "github.com/aixgo-dev/aixgo"

func main() {
    aixgo.Run("config/agents.yaml")
}

Where to read further

Quick Start Guide - Get running in 5 minutes
Pattern Catalog - All 13 orchestration patterns
Configuration examples - Working YAML for every agent type, provider and pattern
GitHub Repository - Source code and examples
Full Feature Reference - Complete technical documentation