HappyClawvsShannon
Generated from the two spec rows. Green marks the better value where a spec has a clear direction. Everything else is just different.
HappyClaw
riba2534 · Agent harness
OSSMCP
Panel
6.82 spec wins
- Reliability
- 6.5
- Usefulness
- 6.8
- Cost
- 7.8
- Longevity
- 6.0
“It answers on Feishu, Telegram, QQ, DingTalk, WeChat, WeCom, Discord and WhatsApp, so the agent is now unreachable in eight places at once.”
Shannon
Kocoro Lab · Agent harness
OSSMCP
Panel
7.03 spec wins
- Reliability
- 7.0
- Usefulness
- 6.7
- Cost
- 8.3
- Longevity
- 6.2
“It arrives as a Docker Compose stack with a workflow engine and a policy engine, so your agent framework now needs its own platform team.”
Spec by spec
| Spec | HappyClaw | Shannon |
|---|---|---|
| Architecture | ||
| Category | Agent harness | Agent harness |
| Runs | local, sandbox | local, sandbox |
| Platforms | macos, linux, web | macos, linux |
| Context window | not documented | not documented |
| Protocols | ||
| MCP client | Yes | Yes |
| MCP server | Yes | No |
| Capabilities | ||
| Runs terminal commands | Yes | Yes |
| Multi-file edits | Yes | No |
| Git operations | No | No |
| Browser control | No | No |
| Sandboxed execution | Yes | YesCode execution is isolated in a WASI sandbox inside the Rust agent core, and the whole stack ships as Docker Compose services. |
| Multi-agent orchestration | YesA built-in HappyClaw agent always exists, and users create custom agents such as code review, research or operations through a conversational agent builder. | Yes |
| Headless / CI mode | No | Yes |
| Models | ||
| Backbone | Claude Code | OpenAI, Anthropic, Google, DeepSeek, xAI, Ollama, LM Studio, vLLM |
| Bring your own model | Yes | Yes |
| Local models | No | Yes |
| Cost | ||
| Pricing model | byok | byok |
| Starts at | $0/mo | $0/mo |
| Free tier | Yes | Yes |
| Bring your own key | Yes | Yes |
| Openness | ||
| Open source | Yes | Yes |
| License | MIT | MIT |
| GitHub stars | 835 | 2,279 |
Which one would each critic pick
| Critic | HappyClaw | Shannon | Pick |
|---|---|---|---|
| El Juez | — | — | not enough reviews |
| El Amigo | 6.8 | 7.0 | Shannon — Pick Shannon when agent spend is your actual problem, because budgets are enforced per task; pick VoltAgent if you want something lighter to stand up. |
| El Crítico | 6.3 | 6.5 | Shannon — Generated code executes inside a WASI sandbox, which is a strong boundary and a narrow one: anything needing real filesystem or network access falls outside the guarantee. |
| El Profesor | 7.3 | 7.8 | Shannon — Running agent workflows on a durable execution engine makes every run replayable step by step, which turns debugging from archaeology into reproduction. |
| La Inversora | 6.3 | 6.0 | HappyClaw — 817 stars, a permissive licence and a single maintainer, with distribution that runs entirely through chat platforms rather than through any commercial channel. |
| La Jefa | 6.3 | 7.3 | Shannon — Multi-tenant isolation is enforced by policy rules and destructive steps require human approval, which is the first open project this quarter that anticipated my questions. |
| El Hacker | 8.0 | 7.8 | HappyClaw — MIT, self-hosted in a container, and it is both an MCP client and an MCP server, with the servers scoped per workspace alongside that workspace's own credentials. |
Picks are derived from each critic's own scores. Humans vote on matchups on the duels page.