Skip to main content
skill CX Codex

diff-analyze — Claude Code skill: what it does, how to install

Analyzes a git diff before a PR, scoring risk, classifying changes, and suggesting reviewer roles. Use it before merging a large change.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add ruvnet/ruflo --skill diff-analyze -a claude-code
Open on GitHub

What diff-analyze does

The skill selects a ref for uncommitted work, a branch, or a PR, then runs Ruflo diff-analysis tools. It returns an overall analysis, a quantified risk score, a change category, line and file statistics, plus risk checks for flagged files.

Risk signals include files with high churn, paths related to authentication, cryptography, or permissions, diffs over 500 lines, cross-module edits, and database migration files. With --pr, it resolves the base and head refs through gh pr view N --json baseRefName,headRefName.

Reviewer output describes roles inferred from paths and risk. Check CODEOWNERS separately before assigning people; that matters for any real review handoff.

What it helps with

  • score the risk of a diff before a PR
  • classify changes as features, bug fixes, or refactors
  • suggest reviewer roles from file paths
  • inspect risk in flagged files
  • count changed lines and files

Example prompts for the agent

  • Run diff-analyze for my uncommitted changes.
  • Use diff-analyze --branch BRANCH and assess this diff.
  • Run diff-analyze --pr PR# and suggest reviewer roles.

How to install diff-analyze

1Skills CLI (Vercel)

npx skills add ruvnet/ruflo --skill diff-analyze -a claude-code

2Manually: put the files in place

~/.claude/skills/diff-analyze/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/diff-analyze/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about diff-analyze

How do I install diff-analyze in Claude Code?

Run npx skills add ruvnet/ruflo --skill diff-analyze -a claude-code, then run npx skills add ruvnet/ruflo --skill diff-analyze -a claude-code again. The skill file is placed at ~/.claude/skills/diff-analyze/SKILL.md.

What does diff-analyze do?

It analyzes git diffs for risk, change type, reviewer roles, and diff statistics. It can also inspect risk at the individual file level.

Does diff-analyze assign reviewers?

It recommends reviewer roles based on paths and risk. Check CODEOWNERS separately when assigning specific people.

Source text

SKILL.md · MIT
---
name: diff-analyze
description: Analyze git diffs for risk scoring, reviewer recommendations, and change classification. Use when preparing a PR, reviewing a large or cross-module change, or before merging to assess risk and pick reviewers.
argument-hint: "[--branch BRANCH] [--pr PR#]"
allowed-tools: mcp__plugin_ruflo-core_ruflo__analyze_diff mcp__plugin_ruflo-core_ruflo__analyze_diff-risk mcp__plugin_ruflo-core_ruflo__analyze_diff-classify mcp__plugin_ruflo-core_ruflo__analyze_diff-reviewers mcp__plugin_ruflo-core_ruflo__analyze_diff-stats mcp__plugin_ruflo-core_ruflo__analyze_file-risk Bash
---

# Diff Analysis

Analyze git diffs for risk, complexity, and reviewer assignment.

## When to use

Before submitting a PR or after making significant changes, analyze the diff to understand risk level, get reviewer recommendations, and classify the type of change.

## Steps

0. **Pick the ref** — `HEAD` for uncommitted work, `main..BRANCH` for a branch (`--branch`), or the PR's base..head range (`--pr`, resolve it with `gh pr view N --json baseRefName,headRefName`). The analyze_* tools take a `ref` and run git themselves; they do not accept diff text.
1. **Analyze diff** — call `mcp__plugin_ruflo-core_ruflo__analyze_diff` with `{ ref, includeFileRisks: true }` for a comprehensive analysis
2. **Risk score** — call `mcp__plugin_ruflo-core_ruflo__analyze_diff-risk` with `{ ref }` for a quantified risk assessment
3. **Classify changes** — call `mcp__plugin_ruflo-core_ruflo__analyze_diff-classify` with `{ ref }` to categorize (feature, bugfix, refactor, etc.)
4. **Get reviewer roles** — call `mcp__plugin_ruflo-core_ruflo__analyze_diff-reviewers` with `{ ref, limit }` for roles inferred from paths and risk; check CODEOWNERS separately when assigning people
5. **Diff stats** — call `mcp__plugin_ruflo-core_ruflo__analyze_diff-stats` with `{ ref }` for line counts, file counts, complexity metrics
6. **File-level risk** — call `mcp__plugin_ruflo-core_ruflo__analyze_file-risk` with `{ path }` (required) for each file the earlier steps flagged

## Risk factors

- Files with high churn history
- Security-sensitive paths (auth, crypto, permissions)
- Large diffs (>500 lines)
- Cross-module changes
- Database migration files

© ruvnet · MIT · ruvnet/ruflo

Similar skills

Everything for Code review and refactoring →

From the same set146

cognitive-pattern Defines reasoning patterns for agents and turns them into executable workflows. Use it for decision trees, task decomposition, analogy-based reasoning, or multi-perspective choices. CX Codex +1 756installs ruflo-doctor Runs a health check on a Ruflo install: Node.js, npm, git, config, daemon, memory database, API keys, MCP servers. Add --fix and it repairs the usual breakages. CX Codex +1 755installs test-gaps Detects missing test coverage and suggests tests to write. Helpful right after adding a new feature or before cutting a release. CX Codex +1 753installs autopilot-loop A skill that runs one iteration of an autonomous loop: checks progress, picks the next task, schedules the next wake. Use it when you want Claude to keep working without manual breaks between tasks. CX Codex +1 752installs research-synthesize Turns findings stored in ruflo memory into one report, with evidence grades, contradiction handling and next steps. Useful after a run of deep-research. CX Codex +1 751installs vector-search Runs vector search over corpora via HNSW embeddings and routes queries across up to 11 hot patterns in WASM. CX Codex +1 751installs memory-search Searches Ruflo memory using semantic retrieval or hybrid keyword matching. It pulls task context and past solutions directly into the agent. CX Codex +1 750installs ddd-aggregate Scaffolds a DDD aggregate root with an entity, value objects, a repository interface, domain events, and test stubs. Useful when modeling business logic in an existing bounded context. CX Codex +1 749installs

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor