Stats
Actions
Tags
Help us improve
Share bugs, ideas, or general feedback.
From quorum
Run a single quorum agent (claude, codex, copilot, cursor, or gemini) with a prompt. Usage - /quorum:agent <agent> <prompt>
npx claudepluginhub nexusentis/quorum --plugin quorumHow this skill is triggered — by the user, by Claude, or both
Slash command
/quorum:agentThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Test a single quorum agent by sending it a prompt and returning the raw result.
Spawns external CLI AIs (claude, cursor-agent, codex, gemini) as isolated sub-agents with dedicated context. Delegates tasks to specialized agents via natural language.
Guides Claude Code subagent development: agent files, YAML frontmatter, tool/model config, lifecycle/resumption, CLI/SDK usage, priority, built-in agents, troubleshooting via docs delegation.
Guides creation and configuration of autonomous agents for Claude Code plugins, covering frontmatter, triggering descriptions, system prompts, tools, teams, permissions, and best practices.
Share bugs, ideas, or general feedback.
Test a single quorum agent by sending it a prompt and returning the raw result.
/quorum:agent <agent> <prompt>
Where <agent> is one of: claude, codex, copilot, cursor, gemini, or all (to test all enabled agents).
Examples:
/quorum:agent copilot say hello/quorum:agent cursor what files are in the repo?/quorum:agent all say helloReply with only the word 'ok'Determine the absolute path of the current project root (the directory the user is working in).
Prepend to the prompt:
You are operating in READ-ONLY analysis mode. You may read files and explore the codebase, but do NOT write, modify, or delete any files.
Only produce analysis or recommendations as plain text.
External agents (codex, copilot, cursor, gemini): Call the MCP tool directly with the workdir parameter:
codex → call mcp__quorum-codex__codex_query with prompt and workdircopilot → call mcp__quorum-copilot__copilot_query with prompt and workdircursor → call mcp__quorum-cursor__cursor_query with prompt and workdirgemini → call mcp__quorum-gemini__gemini_query with prompt and workdirClaude agent: Spawn via Task tool as quorum:claude-agent.
Call mcp__quorum__quorum_query with the prompt and workdir to fan out to all external agents in parallel, and spawn quorum:claude-agent via Task tool simultaneously. Skip reconciliation — just report raw results.
For each agent, report:
### <Agent> — <status> (<latency>)
<response or error message>
Keep it simple. No judge layer, no reconciliation. This is a diagnostic tool.