multi

One-shot setup wizard — detects CLIs, configures Exa + Context7 MCPs on each

Delegate execution of a specific plan or plan step to Codex. Use for rigorous implementation on a well-defined task with logic, math, or high detail. Distinct from codex-rescue (open-ended) — use this when the plan is clear.

Use when the user asks Copilot to design an implementation plan, outline steps, or produce a strategy before coding. Invokes Copilot's /plan slash command (Plan Mode) for codebase-aware planning grounded in GitHub context.

Use when the user asks for research that benefits from GitHub context — repo search, issue history, code hosted on GitHub. Invokes Copilot's /research slash command for deep investigation across GitHub and the web.

Use when the user asks for a code review with GitHub context — recent changes, PR-style feedback, or review grounded in repo history. Invokes Copilot's /review code review agent.

Hypothesis-driven root-cause debugging via Cursor's Debug mode. Use when a bug is hard to reproduce or understand, or when log-based investigation would help isolate the issue.

Use when the user asks Cursor to design an approach, outline steps, or produce a plan before coding. Cursor runs in Plan mode (read-only — no file edits).

Bulk or multi-file code writing. Use when the main context shouldn't absorb large diffs, or when the task is clearly a pattern-following implementation across many files.

Fast codebase exploration with Gemini 3 Flash — ingest large amounts of code across many files, answer structural questions, trace call paths, and summarize unfamiliar codebases using Gemini's 1M-token context. Read-only. Use when Claude needs broad, quick codebase orientation without burning main-thread context on file reads.

Deep external research with Gemini 3.1 Pro — web search (Exa), Context7 library docs, and synthesis of outside knowledge into informed design choices. Read-only. Use when Claude needs to investigate APIs, libraries, best practices, or external specifications and fold the findings into a design or recommendation without burning main-thread context.

Use when the user asks to delegate coding, refactors, or multi-file edits to Qwen Code. Good for tasks where Qwen3-Coder's strengths apply or when the main context shouldn't absorb large diffs.

Rewire which CLI handles which role in cc-multi-cli-plugin, OR diagnose/work around an upstream CLI quirk via env vars and config files. Use when the user asks to swap CLIs, change a subagent's target CLI, add or disable a subagent or command, restrict a CLI, hardcode a model, or modify a role's prompt template — and also when a CLI is misbehaving (hangs, missing tools, broken release) and the user needs operator escape hatches like CURSOR_AGENT_PATH, ACP_TRACE, or per-CLI allowlist/MCP config tuning. Works for any CLI in the marketplace — the four default CLIs (Codex, Gemini, Cursor, Copilot) and any additional CLIs the user added via the multi-cli-anything skill. Trigger phrases include "swap Gemini and Cursor", "make Copilot the writer", "disable cursor-debugger", "restrict Codex to read-only", "change which CLI handles debugging", "add /<cli>:<command>", "only install the plugins I need", "hardcode a model for <some-role>", "use /plan from copilot for planning", "cursor is hanging / broken / stuck", "pin an older cursor build", "see what ACP traffic the CLI is sending".

Add a new CLI provider to cc-multi-cli-plugin (beyond the built-in Codex/Gemini/Cursor/Copilot). Use when the user asks to integrate another AI CLI like Qwen, OpenCode, Aider, or any CLI that speaks ACP, ASP, or another structured protocol. Trigger phrases include "add Qwen to the plugin", "integrate OpenCode", "hook up my custom CLI", "support another model via ACP".

3 hooks across 3 events