Token导航 LogoToken导航TokenDH.com
开发执行命令clawhub未标认证来源可访问clear审计提醒

shadows-smart-commit阴影智能提交

Agent Skill

shadows-smart-commit 用于处理 GitHub 仓库、Issue、Pull Request 和代码协作信息,适合在 OpenClaw 中需要围绕仓库状态、代码变更或协作事项进行整理时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

8,397

周安装

343

GitHub Stars

公开资料未说明

下载量

2,717
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:shadows-smart-commit(阴影智能提交)
来源仓库:https://github.com/nakedoshadow/shadows-smart-commit
安装命令:
openclaw skills install shadows-smart-commit
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install shadows-smart-commit

简介

智能 git 提交助手,支持差异分析、常规提交强制与秘密检测。

  • 适合在代码提交时生成有意义消息并防止敏感信息泄露。
  • 通过 clawhub 安装并使用 openclaw skills install shadows-smart-commit 命令启用。
  • 需确认其是否具备对本地仓库的操作权限及回滚能力。
  • 建议结合 Git 工作流评估其对团队协作的影响。

SKILL.md

name
smart-commit
description
Intelligent git commit assistant — analyzes diffs, enforces conventional commits, detects secrets, generates meaningful messages. Use when committing code changes.
metadata
{ "openclaw": { "emoji": "🔒", "homepage": "https://clawhub.ai/NakedoShadow", "requires": { "bins": ["git"] }, "os": ["darwin", "linux", "win32"] } }

Smart Commit — Intelligent Git Commit Assistant

Version: 1.1.0 | Author: Shadows Company | License: MIT


WHEN TO TRIGGER

  • User says "commit", "smart commit", "save my changes", or "push"
  • After completing a coding task when changes need to be committed
  • User asks to review and commit staged changes

WHEN NOT TO TRIGGER

  • User explicitly says "don't commit" or "no commit"
  • No changes exist in the working tree (git status shows clean)
  • User is just browsing or reading code

PREREQUISITES

This skill requires only git on PATH. All commands are standard git operations that read repository state and create commits. No network access, no external APIs, no additional toolchains required.


PROCESS

Phase 1 — Reconnaissance

Run these commands to understand the current state:

git status
git diff --cached --stat
git diff --stat
git log --oneline -5

Analyze:

  1. What files changed (staged vs unstaged)
  2. The nature of changes (new feature, bugfix, refactor, docs, test, chore)
  3. Recent commit style to match existing conventions

Phase 2 — Security Scan (MANDATORY)

Before ANY commit, scan staged changes for leaked secrets:

# Scan staged diff for secret patterns
git diff --cached | grep -inE "(api[_-]?key|secret|token|password|credential|private[_-]?key)\s*[:=]\s*['\"][^'\"]{8,}" || echo "PASS: No secrets detected in staged changes"

If secrets detected: STOP immediately. Warn the user with the exact file:line. Do NOT proceed with the commit.

Check for files that should never be committed:

# Check if dangerous files are staged
git diff --cached --name-only | grep -iE "\.(env|pem|key|p12|pfx)$|credentials|secret" || echo "PASS: No sensitive files staged"

Blocked file patterns: .env, .env.*, *.pem, *.key, *.p12, *credentials*, *secret*, node_modules/, __pycache__/, .DS_Store

Limitations: This grep-based scan catches common patterns but may produce false positives (e.g., test fixtures with "password" in variable names) or miss obfuscated secrets. For high-security projects, complement with gitleaks or trufflehog.

Phase 3 — Staging Strategy

  • NEVER use git add . or git add -A — these can accidentally include sensitive files
  • Stage files individually by name: git add src/feature.ts tests/feature.test.ts
  • Group related files logically in the same commit
  • If unrelated changes exist, suggest splitting into multiple atomic commits

Phase 4 — Commit Message Generation

Follow Conventional Commits specification (conventionalcommits.org):

type(scope): concise description

[optional body explaining WHY, not WHAT]

Co-Authored-By: Claude <noreply@anthropic.com>

Types: feat, fix, refactor, docs, test, chore, perf, style, build, ci

Rules:

  • Subject line: max 72 characters, imperative mood ("add" not "added")
  • Focus on WHY, not WHAT (the diff shows WHAT)
  • Include scope when changes are localized to a module
  • Use HEREDOC for multi-line messages to preserve formatting:
  git commit -m "$(cat <<'EOF'
  type(scope): subject line

  Body explaining the motivation.

  Co-Authored-By: Claude <noreply@anthropic.com>
  EOF
  )"

Phase 5 — Verification

After commit, verify success:

git log --oneline -1
git status

Confirm: commit SHA visible, working tree status as expected.


SECURITY CONSIDERATIONS

  1. Read-only analysis: Phases 1-2 only read git state (status, diff, log). No files are modified, no network calls are made.
  1. Secret detection output: Phase 2 may display matched secret-like patterns in terminal output. Run in a secure terminal where output is not forwarded to shared logging systems.
  1. Write operations: Phase 3 (git add) and Phase 4 (git commit) modify git state. These are local operations — no data is pushed to any remote unless the user explicitly requests git push afterward.
  1. No persistence: This skill does not store credentials, modify config files, or install packages. Each invocation is stateless.
  1. No network access: The entire workflow is local. git push is never executed unless the user explicitly requests it as a separate step.

OUTPUT FORMAT

## Changes Detected
- [file list with change type: added/modified/deleted]

## Security Scan
- Secrets in diff: [PASS — none detected / FAIL — found at file:line]
- Sensitive files: [PASS — none staged / FAIL — list of files]

## Proposed Commit
type(scope): message

## Files to Stage
- [explicit file list, one per line]

## Post-Commit
- SHA: [short hash]
- Status: [clean / remaining unstaged changes]

RULES (by priority)

  1. Security first — NEVER commit secrets, tokens, API keys
  2. Specific staging — NEVER git add . — always name files explicitly
  3. Conventional formattype(scope): message always
  4. Meaningful messages — explain intent, not just what changed
  5. No force push — NEVER git push --force on main/master
  6. No amend — create NEW commits unless explicitly asked to amend
  7. No skip hooks — NEVER use --no-verify
  8. Atomic commits — one logical change per commit

Published by Shadows Company — "We work in the shadows to serve the Light."

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

83.89%
按下载量换算2,279

安全审计

VirusTotal

通过

ClawScan

可疑

Static analysis

通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 openclaw skills install shadows-smart-commit 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills