Send the active Claude plan to other AI models (Codex 5.3, Gemini 3.1 Pro) via OpenRouter for a second opinion.
How this skill is triggered — by the user, by Claude, or both
Slash command
/secondpass-multireview:secondpass_multireviewThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Get a second opinion on your current plan from other AI models via OpenRouter.
Get a second opinion on your current plan from other AI models via OpenRouter.
Gather context — Build a comprehensive review payload with THREE sections. All three are required — do not skip the diff or code context.
Section 1: Plan summary — Summarize from the conversation:
Section 2: Code diff — Get the actual changes:
git diff HEAD~1 (or appropriate range)git diff and/or git diff --cachedSection 3: Relevant code context — Read and include surrounding code that reviewers need to understand the change:
# ... (trimmed) for long sectionsLocate the script — Find the review script. It ships with this plugin:
SCRIPT=$(find ~/.claude/plugins -path "*/secondpass-multireview/scripts/secondpass_multireview.py" 2>/dev/null | head -1)
If not found in plugins, check the project directory:
SCRIPT=${SCRIPT:-$(find . -path "*/secondpass_multireview.py" 2>/dev/null | head -1)}
Write a temp file and send — Assemble all sections into a temp file, then pipe to the script:
source ~/.zshrc 2>/dev/null && python3 "$SCRIPT" < /tmp/secondpass_review.md
Format the temp file (/tmp/secondpass_review.md) as:
## Plan
<plan summary>
## Diff
```diff
<git diff output>
<relevant surrounding code>
The script queries these models in parallel via OpenRouter:
- **GPT-5.3 Codex** (OpenAI)
- **Gemini 3.1 Pro** (Google)
Override models with: `SECONDPASS_MODELS="model/a,model/b" python3 "$SCRIPT"`
Present results — Show each model's feedback clearly, organized by model. Highlight any:
Summarize — End with a brief synthesis: what the models agree on, key concerns raised, and whether the plan should be adjusted.
OPENROUTER_API_KEY environment variable must be setCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub stopman/secondpass-multireview --plugin secondpass-multireview