跳到正文
cccskills
免费在 GitHub 公开

nw-hotspot

Git change frequency hotspot analysis — find the most-changed files in your codebase

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(1)

  • SKILL.md4.7 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

NW-HOTSPOT: Code Crime Scene Hotspot Lens

Wave: CROSS_WAVE Execution: Inline (no agent — Claude executes git commands directly) Inspiration: Adam Tornhill's "Your Code as a Crime Scene"

Overview

Analyze git change frequency to identify the most-changed files in a codebase. Pure git churn — no complexity metrics. Use as a pre-filter to scope analysis or a post-filter to prioritize existing findings.

How It Works

Run git log --name-only over the configured time period, count commits per file, rank by frequency.

Modes

1. Analyze (default)

Show the most-changed files in the repo. Output is inline only — no files persisted.

/nw-hotspot
/nw-hotspot --top=20
/nw-hotspot --since=12m

Steps:

  1. Run: git log --since={since} --format=format: --name-only | sort | uniq -c | sort -rn
  2. Exclude deleted files (verify each file still exists with test -f)
  3. For each file, get last changed date: git log -1 --format="%ar" -- {file}
  4. Display ranked table (top N):
 Rank │ Commits │ Last Changed │ File
──────┼─────────┼──────────────┼─────────────────────────
    1 │      87 │ 2 days ago   │ src/services/payment.ts
    2 │      64 │ 1 week ago   │ src/api/controllers/user.ts
    3 │      51 │ 3 days ago   │ src/models/order.ts
  1. Show summary: total files changed in period, time period, top N shown

Defaults: --top: 15, --since: 6 months

2. Rank (post-filter)

Overlay churn data on an existing analysis report to prioritize findings.

/nw-hotspot --rank code-smell-detector-report.md

Steps:

  1. Run the same git churn analysis
  2. Read the specified report file
  3. Extract file paths mentioned in the report
  4. Annotate each file with its commit count
  5. Re-sort findings by churn (highest first). Findings without extractable file paths go at the bottom.
  6. Show noise reduction: "X of Y findings are in low-churn files (< 20 commits)"

3. Detail (deep-dive)

Deep-dive into a single file's change history.

/nw-hotspot --detail src/services/payment.ts

Steps:

  1. Monthly commit breakdown over the period
  2. Top contributors to the file
  3. Co-change coupling: top 5 files that frequently change alongside this one
  4. Last 10 commit messages for context

4. JSON export (pre-filter composition)

/nw-hotspot --top=10 --json

Output: [{"path": "src/services/payment.ts", "commits": 87, "last_changed": "2 days ago"}, ...]

Parameters

FlagDefaultDescription
--top15Number of files to show
--since6mTime period (e.g., 3m, 12m, 2024-01-01)
--jsonfalseOutput as JSON array
--rank-Path to existing report to re-rank by churn
--detail-Path to single file for deep-dive

Edge Cases

  • No commits in period: Display "No commits found since {date}" rather than an empty table
  • File renames: Counted as separate files (no --follow; keeps it simple and fast)
  • Empty repo / not a git repo: Display error and exit

Usage from Other Skills

Pre-filter (scope downstream analysis)

Run /nw-hotspot --top=10 --json first, then pass the file list to:

  • code-smell-detector: "Analyze only these files: {hotspot list}"
  • cognitive-load-analyzer: "Focus analysis on these files: {hotspot list}"
  • refactoring-expert: "Prioritize recommendations for these files"

Post-filter (rank existing findings)

Run /nw-hotspot --rank {report.md} after any analysis to prioritize by churn.

In nWave workflows

  • Before /nw-refactor — identify which files to refactor first
  • Before /nw-review — focus review effort on high-churn areas
  • Before /nw-root-why — check if the problematic area is a known hotspot

Next Wave

Usage Context: Inline utility, composable with any analysis skill Handoff To: Invoking context (no sequential handoff)

Examples

Example 1: Quick scan of a legacy codebase

/nw-hotspot

Example 2: Scope a code smell analysis to hotspots

/nw-hotspot --top=10 --json

Example 3: Prioritize an existing refactoring report

/nw-hotspot --rank code-refactoring-report.md

Example 4: Investigate a specific troubled file

/nw-hotspot --detail src/services/payment.ts

Expected Outputs

No persistent files. Output is displayed inline. JSON mode outputs a JSON array for composition with other tools.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation

中文简介正在准备中。当前显示原文说明。

nWave-ai/nWave6222026年9月16日 更新

Review dimensions for validating agent quality - template compliance, safety, testing, and priority validation

中文简介正在准备中。当前显示原文说明。

nWave-ai/nWave6222026年9月16日 更新

Review dimensions for acceptance test quality - happy path bias, GWT compliance, business language purity, coverage completeness, walking skeleton user-centricity, priority validation, observable behavior assertions, traceability coverage, and walking skeleton boundary proof

中文简介正在准备中。当前显示原文说明。

nWave-ai/nWave6222026年9月16日 更新

Detailed 5-phase workflow for creating agents - from requirements analysis through validation and iterative refinement

中文简介正在准备中。当前显示原文说明。

nWave-ai/nWave6222026年9月16日 更新

5-layer testing approach for agent validation including adversarial testing, security validation, and prompt injection resistance

中文简介正在准备中。当前显示原文说明。

nWave-ai/nWave6222026年9月16日 更新

Architectural style selection decision matrices, trade-off analysis, structural enforcement rules, and combination patterns. Load when choosing or evaluating architecture styles.

中文简介正在准备中。当前显示原文说明。

nWave-ai/nWave6222026年9月16日 更新

查看 nWave-ai 的全部技能

报告此技能的问题