Skill Details
Runtime
(claude-4-8-opus)
Example user message
[userMessage]: turn this prompt into a skill: "You are a senior code reviewer. Review the [LANGUAGE] code in [FILE] for security, performance, and style. Output findings as a markdown table sorted by severity, max 20 rows."
Example agent response
Built the skill. Mapped the [LANGUAGE] and [FILE] slots into the description and Process as free-form inputs (no Variables section), and lifted the "markdown table, max 20 rows, sorted by severity" rules into Validation checkboxes.
---
name: code-review-reporter
description: Reviews source code for security, performance, and style and outputs a severity-sorted findings table. Use when a developer says "review this code", "audit this file for issues", or "give me a code review table".
---
# Code Review Reporter
## Purpose ...
## When to Use (6–8 scenarios + SCOPE) ...
## Process (1. Ingest file → ... → emit table) ...
## Validation (☐ table is markdown ☐ ≤20 rows ☐ sorted by severity) ...
Report: name=code-review-reporter · ~480 tokens · install at ~/.claude/skills/code-review-reporter/SKILL.md
Instant access
Commercial use
Money‑backBy purchasing this skill, you agree to our terms of service
3
Favorites
10
Views
CLAUDE-4-8-OPUS
Stop letting proven prompts rot in scratch files. This skill maps any tested prompt or template into a deployment-ready SKILL.md: it extracts intent, folds variable slots into the description and process, derives activation triggers, lifts hidden guardrails into a Validation checklist, and writes a SCOPE block so the skill fires on the right intents — not everything. Output is a single canonical SKILL.md plus a build report with name, token estimate, and install path.
...more
Added 8 hours ago
