Review Answers — Development skill for Claude Code
Host one turn of the stage-(7) review conversation — answer the user's CR comments from the artifacts, dispatch a seat when judgment is needed, and act only on what they confirm.
How to install Review Answers
Installs to ~/.claude/skills/taikerliang-roz-gate-review-answers/SKILL.md
mkdir -p ~/.claude/skills/taikerliang-roz-gate-review-answers && curl -fsSL https://raw.githubusercontent.com/TaikerLiang/roz-gate/HEAD/commands/review-answers.md -o ~/.claude/skills/taikerliang-roz-gate-review-answers/SKILL.md Restart Claude Code, or start a new session, for it to be picked up.
What Review Answers does
description: Host one turn of the stage-(7) review conversation — answer the user's CR comments from the artifacts, dispatch a seat when judgment is needed, and act only on what they confirm argument-hint: "[issue-number]"
Host **one turn** of the user's review at stage (7) (see `${CLAUDE_PLUGIN_ROOT}/references/workflow.md` → (7) Review). The user is reading the CR and talking; you answer, and you act only on what they confirm. Follow these steps; do nothing beyond them.
0. Load co
Alternatives in Development
- Domxss — Confirm DOM XSS in a real headless browser — injects canary payloads into params + URL fragment and only repor 4.5k ★
- Discuss — Act as a planning interviewer 2.9k ★
- Remove Comments — Remove Obvious Comments 2.8k ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
Blast Radius Bench
A benchmark for agentic coding-tool judgment under ambiguity: does the agent confirm before touching ambiguous
Planwise
Planwise is a Claude Code plugin that brings end-to-end project management to your CLI. Create plans, execute
Founder Voice Reply
Reply to tweets, posts and comments in the founder's voice. Loads context/voice-profile.md, finds the angle on
Respond Reviewer
Classify a reviewer letter into per-comment labels and dispatch each to the right specialist; assemble a rebut
Add Backlog Item
Backlog stage. Creates one backlog work item from natural-language input with tracker destination detection an
Openprism Bakeoff
Blind-judge two models on one prompt to decide which earns a panel seat
Related Agents
Braintrust Analyst Agent
You are a specialized analysis agent. Your job is to run Braintrust analysis scripts, interpret results, and w
X Codebase Explorer
Read-only codebase tracer. Answers where something lives, what depends on it, and what a change would touch, w
Feature Architect
Use to turn a vague client ask into a buildable feature spec — scoping, decisions-needed identification, skill