Multisage

Query Multisage for multi-expert AI answers from Claude, GPT, and Gemini. Use this skill whenever the user wants multiple AI perspectives on a question, says...

This skill appears to be what it claims (a wrapper around a Multisage CLI) but the package metadata in the registry is incomplete: SKILL.md clearly requires installing 'multisage' via npm and setting MULTISAGE_API_KEY, yet the registry lists no required env vars or installs. Before installing or using it: 1) Verify the official npm package name and publisher on npmjs.com and confirm it's the legitimate multisage package. 2) Prefer creating a restricted API key (if supported) rather than using a full-account key. 3) Avoid blindly running 'npm install -g' from unknown sources; consider installing in a sandbox or container. 4) Be cautious about commands that grep dotfiles (.env, ~/.env) or echo partial keys — these touch sensitive data. 5) Check Multisage's privacy/security documentation (rates, what content is sent to downstream providers) and confirm you are comfortable with the CLI sending prompts to external services. If possible, ask the skill author or registry maintainer to update the metadata to declare the required env var and any install steps so the registry and SKILL.md match.

Type: OpenClaw Skill Name: multisage Version: 2.2.0 The multisage skill bundle contains instructions in SKILL.md that direct the AI agent to perform high-risk operations, including installing a global NPM package and searching for API keys within sensitive files like ~/.env (credential hunting). While these actions are contextually related to the tool's setup, they represent significant security risks when executed by an automated agent. Additionally, the _meta.json file contains an unusual future-dated timestamp (2026). No clear evidence of malicious exfiltration was found, but the broad file access and installation commands are inherently risky.