#7171st of 80 in Agent platforms
shannon
Multi-agent orchestration on Temporal with token budgets and replay debugging
- Stars
- 2.3k
- License
- MIT
- Last commit
- Sep 2026
- Last release
- Jun 2026
- Language
- Go
1,696 on 21 Apr 20262,327 on 10 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Shannon is a multi-agent orchestration stack with a Go gateway and orchestrator, a Rust agent core with a WASI sandbox, and a Python LLM service. Tasks are submitted over REST, an OpenAI-compatible API or a Python SDK, then routed by complexity to strategies such as DAG, ReAct, research and swarm. Temporal provides workflow replay, and per-task token budgets fall back to cheaper models.
Who it is for: Teams self-hosting production agent workflows with cost controls
Strengths
- Temporal workflows allow step-by-step replay of failed runs via scripts/replay_workflow.sh
- Hard per-task token budgets via budget_max, with fallback to cheaper models
- Code runs in a WASI sandbox; OPA policies can gate actions behind human approval
- OpenAI-compatible endpoint plus REST, SSE, WebSocket and a Python SDK
Weaknesses
- Many moving parts: gateway, orchestrator, agent core, LLM service, Temporal, PostgreSQL, Redis
- Needs an external LLM API key or a separately hosted local model endpoint
- Playwright browser image is 3.4GB and is not started by default
- Source build needs Go, Rust and Python toolchains; quick install is a curl-pipe-bash script
What it needs
- no GPU
- Compose
- Needs Docker Compose, Temporal, PostgreSQL, Redis, LLM provider API key
- Models: OpenAI, Anthropic, Google Gemini, xAI, DeepSeek
- port 8080
Quick answers
Is shannon open source?
Yes. Its code is on GitHub under the MIT license.
Can you run shannon with Docker?
Yes. Its repository has a Docker setup: a Dockerfile, a Compose file or a published image.
Does shannon need a GPU?
No, it runs without one.
Also in Agent platforms
See all 80| Rank | Project | Score |
|---|---|---|
| 1 | AutoGPTRole-based AI agents coordinated by one lead agent, plus a visual workflow builder | 95 out of 100 |
| 2 | daily_stock_analysisLLM-driven daily analysis of watchlist stocks across six markets, pushed to chat apps | 95 out of 100 |
| 3 | LangflowVisual builder for AI agents and workflows, exposed as API or MCP server | 94 out of 100 |
| #4 | deer-flowAgent harness running sub-agents, memory and sandboxes with pluggable skills | 93 out of 100 |
| #5 | DifyVisual builder for LLM workflows, RAG pipelines and agents, with APIs | 92 out of 100 |