second-opinion

Invoke the skill and choose a review tool, scope, and focus. It pipes your git diff to Codex CLI or Gemini CLI in headless mode and returns structured findings with file and line references.

Codex exec produces structured JSON findings with priorities and confidence scores using an output schema, while Gemini uses code-review and security extensions with freeform output. Running both in parallel lets you compare where the models agree.

Yes. The skill supports three scopes: uncommitted changes including untracked files, branch diffs against the auto-detected default branch, and a specific commit via git diff sha~1..sha.

The dependency scan only runs when the diff touches dependency manifest files like package.json, requirements.txt, or go.mod. The scan analyzes the entire dependency tree, so it is skipped when dependencies were not changed.

The skill runs the tool directly and reports the install command if it fails with command not found: npm i -g @openai/codex or npm i -g @google/gemini-cli. If both were selected, it runs only the available one.

Avoid it when neither CLI is installed or authenticated, when reviewing non-code files like documentation, or when you only want the current assistant's own review. Very large diffs over 2000 lines may also need narrowing.