Claw Multi-Agent Skill - Setup Guide¶

Source: zcyynl/claw-multi-agent Skill family: zcyynl/claw-multi-agent (1 installable skill) Combined Installs: ~168 across indexed listings (Sep 29, 2026 snapshot) Category: Automation Quality Tier: 🟡 Unverified (no skills.sh Trust Hub / Socket / Snyk verdicts published - verified Sep 29, 2026)

A single OpenClaw skill published by an individual maintainer that assembles an AI sub-agent squad - researcher, analyst, writer - with different roles and models working in parallel. Per the README (fetched Sep 29, 2026; Chinese, translated here), it offers three modes: commander (parallel web-search agents aggregated into one report), pipeline (lightweight tool-free agents for multi-angle analysis), and hybrid (research then multi-draft output), with auto-routing between modes. The sweep verified it 1/1 with 168 installs.

Installation¶

npx skills add zcyynl/claw-multi-agent

CorpusIQ Use Cases¶

Use Case	How
Parallel research sweeps	Run commander mode to search several topics at once and merge results into one report.
Multi-angle analysis	Use pipeline mode to analyze one problem from technical, product, and beginner perspectives.
Multi-draft reports	Generate several draft variants in hybrid mode after a research phase.

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects

Limitations / Verification¶

Verified: skills.sh indexing of Sep 29, 2026 (1/1; 168 installs).
Not verified: no live install test performed. Star count not captured in the sweep. The README documents an alternative install: npx clawhub@latest install claw-multi-agent, with zero-config reuse of existing OpenClaw models.

Security¶

Audit	Verdict
Gen Agent Trust Hub	Not published
Socket	Not published
Snyk	Not published

Skills Marketplace

← Skills Catalog | Marketplace → Powered by CorpusIQ