Mumo
PassAudited by ClawScan on May 6, 2026.
Overview
Mumo appears to be a coherent instruction-only integration for a paid external deliberation service, but users should knowingly allow API-key use, credit consumption, and sharing prompts or specs with external models.
Install this skill if you want OpenClaw to use mumo for external multi-model review. Before use, verify the mumo.chat endpoint, protect the API key, monitor spending, and avoid sending confidential specs, secrets, credentials, customer data, or proprietary code unless you are comfortable with mumo's data-handling terms.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Anyone with access to that configuration file may be able to use the mumo account key, and agent calls can consume account credits.
The skill requires a mumo bearer API key stored in the local OpenClaw configuration so the agent can call the mumo MCP server.
OpenClaw stores outbound MCP servers in `~/.openclaw/openclaw.json` ... "Authorization": "Bearer mmo_live_YOUR_KEY_HERE"
Use a dedicated mumo API key, keep the OpenClaw config file private, monitor credit usage, and revoke or rotate the key if it is no longer needed.
Private specs, source excerpts, security designs, or business plans included in prompts may leave the local OpenClaw environment.
The skill's core function sends the user's deliberation prompt or referenced design material to mumo and a panel of external model providers.
mumo runs a panel of frontier models in parallel — Claude, GPT, Gemini, Grok, Qwen, Kimi, GLM
Avoid sending secrets, credentials, customer data, or highly confidential material unless mumo's privacy, retention, and provider-sharing terms are acceptable for that data.
Normal use can spend mumo credits, especially if the agent appends follow-up rounds.
The agent is taught to use a paid external deliberation tool and not to check credit balance before every call.
`get_credit` is useful when balance is uncertain ... Don't make cost preflight a default ritual before every deliberation ... Frontier panels typically cost $0.20-0.60 per round.
Set budget expectations up front and ask the agent to confirm before extra rounds or before sending unusually large or sensitive prompts.
If the upstream repository changes, a future install may load instructions different from the artifacts reviewed here.
The manual install path clones a live GitHub repository into the OpenClaw skills directory rather than pinning a reviewed commit or release.
git clone https://github.com/mumo-chat/mumo-openclaw ~/.openclaw/skills/mumo
Install from a trusted release or pinned commit, and inspect the files before enabling the skill.