# Kocoro-lab/Shannon

A production-oriented multi-agent orchestration framework.

Repository: https://github.com/Kocoro-lab/Shannon
Canonical: https://ross.abutalabs.com/products/kocoro-lab-shannon
Homepage: https://shannon.run
Language: Go
License: MIT
License Family: permissive
Topics: agent, ai, multi-agent-systems
Last push: 2026-08-08T02:16:25+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 86, longevity 26
- inputs: {"age_days": 369, "days_push": 26, "days_rel": 92, "gap_med": 12.5, "n_releases_24m": 9}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2216, forks 352 (observed 2026-08-28T04:06:27.180357+00:00)

## What it is
Shannon is a production-oriented multi-agent orchestration framework written in Go (with Rust components) that coordinates AI agent swarms with execution strategies, token budget controls, and human approval workflows. It ships with Temporal-based durable workflows, WASI sandboxing, and built-in observability via event streaming, Prometheus metrics, and OpenTelemetry tracing.

## Use cases
- orchestrate multiple AI agents on complex tasks
- run LLM agents in production with cost controls
- debug agent executions step by step
- enforce token budgets and model fallbacks for agent tasks
- sandbox agent code execution securely
- add human approval gates to automated agent workflows
- avoid vendor lock-in across OpenAI, Anthropic, and local models
- monitor and trace agent systems with Prometheus and OpenTelemetry

## When to choose
- you need production-grade multi-agent orchestration with reliability guarantees
- cost control via hard token budgets is a requirement
- you want durable, replayable agent workflows with time-travel debugging
- security matters: sandboxed execution, OPA policies, multi-tenant isolation
- you want provider-agnostic LLM support including local models via Ollama

## When to avoid
- you only need a simple single-agent chatbot or prompt pipeline
- you want a lightweight Python-native agent library rather than a Docker-based service stack
- your team cannot operate infrastructure like Temporal, Docker Compose, and observability tooling
- you need a minimal embedded SDK inside an existing app rather than a standalone orchestration platform

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, workflow-automation, monitoring, tracing, llm-inference, security, api-framework
- domain: large-language-models, artificial-intelligence, developer-tools, self-hosted
- platform: go, rust, windows, self-hosted, cross-platform
- tags: multi-agent-orchestration, temporal-workflows, token-budgeting, human-in-the-loop, wasi-sandboxing, time-travel-debugging, llm-provider-agnostic, observability, ai-agents, automation, docker, linux, macos

## Member repositories
- Kocoro-lab/Shannon (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:27.180357+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:45:56.268313+00:00, confidence not recorded.
  - readme: https://github.com/Kocoro-lab/Shannon (fetched 2026-08-28T04:06:27.180357+00:00, sha 38f0c4b38878)
  - homepage: https://shannon.run (fetched 2026-08-29T10:26:06.669926+00:00, sha 798f772c0620)
- Data as of 2026-08-30T08:39:29.467469+00:00.