/brainstorm

command: brainstorm
disable-model-invocation: true
description: "Start a creative thought partner brainstorming session"

/octo:brainstorm

INSTRUCTIONS FOR CLAUDE

MANDATORY COMPLIANCE — DO NOT SKIP

When the user invokes `/octo:brainstorm`, reuse an explicitly supplied Solo or Multi-AI mode. Ask the mode question only when that choice is missing or contradictory. Do not infer provider spending approval from an unrelated answer.

---

Step 1: Resolve mode

Read the supplied topic, mode, constraints, and desired output first. Reuse clear answers. Use AskUserQuestion for a material unresolved or contradictory field; when mode is unresolved, ask:

AskUserQuestion({
  questions: [
    {
      question: "How should we brainstorm?",
      header: "Mode",
      multiSelect: false,
      options: [
        {label: "Solo", description: "Claude-only thought partner session — fast and focused"},
        {label: "Team", description: "Multi-AI brainstorm — diverse perspectives from multiple providers"}
      ]
    }
  ]
})

**WAIT for the user's answer before proceeding.**

---

Step 2: Run the Selected Mode

If Solo Mode selected:

Standard thought partner session using four breakthrough techniques:

• Pattern Spotting, Paradox Hunting, Naming the Unnamed, Contrast Creation

**Session flow:** 1. Frame the exploration topic 2. Guided questioning (one question at a time — do NOT dump multiple questions) 3. Challenge generic claims until specific 4. Collaboratively name discovered concepts 5. Export session with breakthroughs summary

**See:** skill-thought-partner for full documentation.

If Team Mode selected:

Step 2a: Display Visual Indicator Banner (MANDATORY)

**You MUST output this banner before doing anything else.** This is NOT optional — users need to see which AI providers are active and understand cost implications.

**MANDATORY: First, use the Bash tool to check provider availability:**

set -euo pipefail

echo "PROVIDER_CHECK_START"
printf "codex:%s\n" "$(command -v codex >/dev/null 2>&1 && echo available || echo missing)"
printf "perplexity:%s\n" "$([ -n "${PERPLEXITY_API_KEY:-}" ] && echo available || echo missing)"
printf "opencode:%s\n" "$(command -v opencode >/dev/null 2>&1 && echo available || echo missing)"
printf "copilot:%s\n" "$(command -v copilot >/dev/null 2>&1 && echo available || echo missing)"
printf "qwen:%s\n" "$(command -v qwen >/dev/null 2>&1 && echo available || echo missing)"
printf "ollama:%s\n" "$(command -v ollama >/dev/null 2>&1 && curl -sf http://localhost:11434/api/tags >/dev/null 2>&1 && echo available || echo missing)"
printf "openrouter:%s\n" "$([ -n "${OPENROUTER_API_KEY:-}" ] && echo available || echo missing)"
printf "agy:%s\n" "$(command -v agy >/dev/null 2>&1 && echo available || echo missing)"
printf "grok:%s\n" "$(command -v grok >/dev/null 2>&1 && { [ -n "${XAI_API_KEY:-}" ] || [ -f "${HOME}/.grok/auth.json" ]; } && echo available || echo missing)"
echo "PROVIDER_CHECK_END"

Then render the provider banner from actual provider checks. Do not hand-write or summarize this banner; run this block and display its output exactly. The output MUST include the Antigravity line even when `agy` is missing.

status_cli() { command -v "${1}" >/dev/null 2>&1 && echo "Available ✓" || echo "Not installed ✗"; }
status_env() { [[ -n "${1:-}" ]] && echo "Configured ✓" || echo "Not configured ✗"; }
codex_status="$(status_cli codex)"
agy_status="$(status_cli agy)"
if command -v grok >/dev/null 2>&1 && { [ -n "${XAI_API_KEY:-}" ] || [ -f "${HOME}/.grok/auth.json" ]; }; then grok_status="Available ✓"; else grok_status="Not installed ✗"; fi
opencode_status="$(status_cli opencode)"
copilot_status="$(status_cli copilot)"
qwen_status="$(status_cli qwen)"
if command -v ollama >/dev/null 2>&1 && curl -sf http://localhost:11434/api/tags >/dev/null 2>&1; then ollama_status="Available ✓"; else ollama_status="Not installed ✗"; fi
perplexity_status="$(status_env "${PERPLEXITY_API_KEY:-}")"
cat <<BANNER
🐙 **CLAUDE OCTOPUS ACTIVATED** — Multi-AI Brainstorm
🔍 Brainstorm: [Topic being explored]

Providers:
🔴 Codex CLI: ${codex_status}
🧭 Antigravity CLI: ${agy_status}
🤖 Grok CLI (xAI): ${grok_status}
🟤 OpenCode: ${opencode_status}
🟢 Copilot CLI: ${copilot_status}
🟠 Qwen CLI: ${qwen_status}
⚫ Ollama: ${ollama_status}
🔵 Claude: Available ✓ — Synthesis, pattern naming, and moderation
🟣 Perplexity: ${perplexity_status}
BANNER

The rendered banner must look like this shape, with ACTUAL statuses:

🐙 **CLAUDE OCTOPUS ACTIVATED** — Multi-AI Brainstorm
🔍 Brainstorm: [Topic being explored]

Providers:
🔴 Codex CLI: [Available ✓ / Not installed ✗]
🧭 Antigravity CLI: [Available ✓ / Not installed ✗]
🤖 Grok CLI (xAI): [Available ✓ / Not installed ✗]
🟤 OpenCode: [Available ✓ / Not installed ✗]
🟢 Copilot CLI: [Available ✓ / Not installed ✗]
🟠 Qwen CLI: [Available ✓ / Not installed ✗]
⚫ Ollama: [Available ✓ / Not installed ✗]
🔵 Claude: Available ✓ — Synthesis, pattern naming, and moderation
🟣 Perplexity: [Configured ✓ / Not configured ✗]

**PROHIBITED: Displaying only "🔵 Claude: Available ✓" without listing all providers.** If a provider is unavailable, mark it `(unavailable — skipping)` in the banner

Step 2b: Frame the Topic

Ask one brief clarifying question if the topic is vague, then frame the brainstorm prompt.

Step 2c: Dispatch Parallel Brainstorm Queries (MANDATORY)

**You MUST dispatch to at least 2 providers.** Do NOT brainstorm solo and call it Team mode.

Launch external providers in parallel through Octopus routing. This block waits for every advisor to finish, often several minutes, so run it with `run_in_background: true`; a foreground 600000 ms timeout stops it after 10 minutes, while the launcher waits up to `OCTOPUS_ADVISOR_WAIT_SECONDS` (default 3600 s). It prints each collected answer before it exits, because `$RUN_DIR` is deleted on exit:

TOPIC="[TOPIC]"
ORC