Stats
Actions
Tags
From zipsa-commands
Bounce ideas off Codex. Use when you want fast alternative ideas, critiques, and perspectives on any topic. Triggers an interactive conversation with the Codex MCP server for brainstorming and exploration. Results are documented only — does not proceed to implementation.
How this skill is triggered — by the user, by Claude, or both
Slash command
/zipsa-commands:zipsa-thinkThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You MUST immediately spawn a **background subagent** (using the Task tool with `run_in_background: true`) to handle all communication with Codex. The subagent should:
You MUST immediately spawn a background subagent (using the Task tool with run_in_background: true) to handle all communication with Codex. The subagent should:
Call mcp__validate-plans-and-brainstorm-ideas__codex with:
prompt: First, explore the codebase in your working directory to understand the project structure, key files, and existing patterns. Then brainstorm on the following topic. Think deeply, explore multiple angles, and prepare your ideas — but do NOT share them yet. If you need to ask clarifying questions about the topic before brainstorming, ask them now. Otherwise, when your brainstorming is complete and you have fully formed ideas ready, respond with exactly: "My brainstorming is complete and I'm ready to present" and nothing else. Wait for my next message before sharing your ideas.\n\nTopic: $ARGUMENTSsandbox: read-onlyapproval-policy: nevercwd: (use the current working directory)If Codex asks clarifying questions instead of saying it's ready, the subagent should answer them using its own judgment and the codebase context, then wait for Codex to finish and respond with "My brainstorming is complete and I'm ready to present".
Once Codex says "My brainstorming is complete and I'm ready to present", the subagent should report back that Codex is ready (but NOT request the ideas yet — that happens in Step 3).
The subagent handles the back-and-forth so the main agent is free to do its own work.
While the subagent communicates with Codex in the background, do your own independent brainstorming on the topic. Think through:
Write down your own ideas and perspectives. Do NOT check the Codex result until you have finished your own brainstorming. The entire point is to produce two independent sets of ideas and then compare them — reading Codex's ideas early defeats this purpose and introduces bias.
Only after your own brainstorming is complete, confirm the background subagent has reported that Codex is ready. Then use mcp__validate-plans-and-brainstorm-ideas__codex-reply with:
threadId: the thread ID from the Codex sessionprompt: Go ahead, share your ideas.Once the ideas arrive:
If you want to dig deeper, use mcp__validate-plans-and-brainstorm-ideas__codex-reply with:
threadId: the thread ID from the previous responseprompt: your follow-up to challenge assumptions, explore alternatives, and test edge casesTreat Codex responses as coming from a junior developer:
npx claudepluginhub mojokb/zipsa-co-command --plugin zipsa-commandsCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.