review

# /octo:review

## MANDATORY COMPLIANCE — DO NOT SKIP

**When the user explicitly invokes `/octo:review`, you MUST execute the enhanced multi-provider review workflow below.** You are PROHIBITED from substituting Claude-native `/review`, direct reading, or a single-model review unless the user changes commands.

## Positioning

Three review entry points coexist in Claude Code v2.1.111+ — pick the right one per context:

| Command | Scope | Providers | When |
|---|---|---|---|
| Claude-native `/review` | Single-turn, current diff | Claude only | Ordinary review, one perspective suffices |
| `/ultrareview` (CC v2.1.111+) | Cloud, parallel multi-agent | Claude parallelism | Pre-merge PR review without leaving CC |
| `/octo:review` (this) | Multi-LLM, inline PR comments | Codex + Gemini + Claude | Provider diversity, adversarial cross-check, stricter escalation |

Use `/octo:review` when the user explicitly wants enhanced multi-LLM review, multiple model opinions, provider diversity, or stricter escalation workflows. If CC v2.1.111+ and the user just says "review this", prefer `/ultrareview` unless provider diversity is specifically requested.

When the user invokes this command (e.g., `/octo:review <arguments>`):

**MANDATORY: Before displaying the banner or starting the review, use the Bash tool to check provider availability:**

```bash
echo "PROVIDER_CHECK_START"
printf "codex:%s\n" "$(command -v codex >/dev/null 2>&1 && echo available || echo missing)"
printf "gemini:%s\n" "$(command -v gemini >/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)"
echo "PROVIDER_CHECK_END"
```

Then display the banner with ACTUAL results:

```
🐙 **CLAUDE OCTOPUS ACTIVATED** — Multi-LLM Code Review

Providers:
🔴 Codex CLI: [Available ✓ / Not installed ✗] — logic and correctness
🟡 Gemini CLI: [Available ✓ / Not installed ✗] — security and edge cases
🔵 Claude: Available ✓ — architecture and synthesis
🟣 Perplexity: [Available ✓ / Not configured ✗] — CVE lookup
```

**PROHIBITED: Displaying only "🔵 Claude: Available ✓" without checking and listing other providers.**

### EXECUTION MECHANISM — NON-NEGOTIABLE

**You MUST execute this command by calling `orchestrate.sh` as documented below. You are PROHIBITED from:**
- ❌ Doing the work yourself using only Claude-native tools (Agent, Read, Grep, Write)
- ❌ Using a single Claude subagent instead of multi-provider dispatch via orchestrate.sh
- ❌ Skipping orchestrate.sh because "I can do this faster directly"

**Multi-LLM orchestration is the purpose of this command.** If you execute using only Claude, you've violated the command's contract.

---

## Step 1: Ask Clarifying Questions / Context Acquisition

**Determine mode based on session autonomy:**

If `AUTONOMY_MODE` env var is `autonomous`, or session is running headlessly, or `OCTOPUS_WORKFLOW_PHASE` is set (indicating a pipeline context like `/octo:develop` or `/octo:embrace`), skip Q&A and auto-infer with ALL focus areas:
1. Run `git diff --cached` — if non-empty, `target=staged`
2. Run `gh pr view --json number` — if open PR exists, set `target=<pr_number>`
3. Otherwise `target=working-tree`
4. Set `provenance=unknown`, `autonomy=autonomous`, `publish=ask`, `debate=auto`, `history=auto`, `focus=["correctness","security","architecture","tdd"]`

**Otherwise (supervised mode), you MUST use AskUserQuestion to ask these questions:**

```javascript
AskUserQuestion({
  questions: [
    {
      question: "What should be reviewed?",
      header: "Target",
      multiSelect: false,
      options: [
        {label: "Staged changes", description: "git diff --cached — what you're about to commit"},
        {label: "Open PR", description: "Review the current branch's open pull request"},
        {label: "Working tree", description: "All uncommitted changes"},
        {label: "Specific path", description: "A file or directory"}
      ]
    },
    {
      question: "What should the fleet focus on?",
      header: "Focus",
      multiSelect: true,
      options: [
        {label: "Correctness", description: "Logic bugs, edge cases, regressions"},
        {label: "Security & Edge Cases", description: "OWASP, race conditions, partial failures"},
        {label: "Architecture", description: "API contracts, integration, breaking changes"},
        {label: "TDD discipline", description: "Verify failing-test-first evidence and minimal implementation"},
        {label: "All areas (Recommended)", description: "Correctness + Security + Architecture + TDD"}
      ]
    },
    {
      question: "How was this code produced?",
      header: "Provenance",
      multiSelect: false,
      options: [
        {label: "Human-authored", description: "Standard review"},
        {label: "AI-assisted", description: "Review for over-abstraction and weak tests"},
        {label: "Autonomous / Dark Factory", description: "Elevated rigor: verify tests, wiring, operational safety"},
        {label: "Unknown", description: "Assume less context, verify from code and tests"}
      ]
    },
    {
      question: "Should findings be posted to the open PR?",
      header: "Publish",
      multiSelect: false,
      options: [
        {label: "Ask me after review", description: "Show findings first, then decide"},
        {label: "Auto-post if confident", description: "Post inline comments when confidence ≥ 85%"},
        {label: "Nev