技能 cc-best-practices

📦

cc-best-practices

Name: cc-best-practices
Author: dboeckli

内容修订 r1 中风险 ⚙️ 外部命令📁 文件系统访问🌐 网络访问

Improve Claude Code Workflows

Unclear prompts and crowded sessions reduce coding quality. This skill provides practical workflows for planning, verification, context control, automation, and independent review.

支持: Claude Codex Code(CC)

📊 71 合格

使用我的智能体安装

将此请求复制给你的智能体，其中包含权威 Skill 页面和清单。

智能体请求

Review the Skillstore skill "cc-best-practices" from https://skillstore.io/skills/dboeckli-cc-best-practices.md and its manifest at https://skillstore.io/api/skills/dboeckli-cc-best-practices/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

你的智能体仍应展示其计划，并按安全策略请求必要确认。

Agent 可读资源

当 AI Agent、爬虫或脚本需要干净上下文，而不是读取完整页面时，请使用这些链接。

Markdown 详情 GET /skills/dboeckli-cc-best-practices.md 已签名 manifest GET /api/skills/dboeckli-cc-best-practices/manifest 已签名 lockfile GET /api/skills/dboeckli-cc-best-practices/lockfile

测试它

正在使用“cc-best-practices”。 Help me ask Claude to fix an intermittent login failure.

预期结果:

Investigate login failures after session timeout. Inspect authentication refresh logic, add a reproducing test, implement the fix, and report test evidence.

正在使用“cc-best-practices”。 My session is long, and Claude repeats the same mistake.

预期结果:

Record the missing constraint and any useful evidence.
Clear the unrelated conversation context.
Start again with the constraint, affected files, expected behavior, and a verification command.

正在使用“cc-best-practices”。 How should my team review a large refactor?

预期结果:

Use one session to implement the approved plan and another fresh session to inspect the diff, edge cases, and verification results.

安全审计

中风险

v1 • 2026/9/26 打开版本化报告

All 43 static alerts are false positives: the script uses fixed local validation commands, while the documentation alerts identify Markdown or benign examples. Two contextual risks remain because examples recommend sending logs to Claude and enabling broad unattended execution without adequate cautions.

已扫描文件

371

分析行数

复核项

已忽略误报

已确认安全问题 (2)

中

SKILL.md:142-145

Potential Sensitive Log Disclosure

The skill recommends piping an error log to Claude without advising users to review or redact secrets and personal data first.

The explicit log piping example can transmit complete log contents to an external model service, and no nearby redaction warning is present.

中

SKILL.md:229-249 references/commands.md:31-38

Overbroad Unattended Automation Guidance

Examples allow editing and wildcard git commits, then recommend auto permission mode without requiring review before persistent changes.

The documented commands explicitly grant broad edit and git commit access and promote uninterrupted auto mode, creating clear unintended-change risk.

风险因素

⚙️ 外部命令 (38)

📁 文件系统访问 (2)

scripts/validate-claude-md.sh:10 scripts/validate-claude-md.sh:8

🌐 网络访问 (1)

SKILL.md:10

审计者: codex

分享与引用此报告

分享版本化评估报告、中性徽章、嵌入卡片和引用。Skillstore 只报告证据，不判断该 Skill 是否安全。

打开版本化报告

复制报告链接

https://skillstore.io/skills/dboeckli-cc-best-practices/audits/1?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown 徽章

[![Skillstore security assessment](https://skillstore.io/badges/skills/dboeckli-cc-best-practices/security.svg)](https://skillstore.io/skills/dboeckli-cc-best-practices?utm_source=security_passport_badge)

HTML 徽章

<a href="https://skillstore.io/skills/dboeckli-cc-best-practices?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/dboeckli-cc-best-practices/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

嵌入卡片

<iframe src="https://skillstore.io/embed/skills/dboeckli-cc-best-practices.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

学术引用 (APA · BibTeX · CFF)

APA 引用

dboeckli. (2026). cc-best-practices security audit report (audit version 1) [Author version unspecified]. Skillstore. https://skillstore.io/skills/dboeckli-cc-best-practices/audits/1

BibTeX 引用

@techreport{dboeckli-dboeckli-cc-best-practices-2026,
  author = {dboeckli},
  title = {cc-best-practices security audit report (audit version 1)},
  institution = {Skillstore},
  year = {2026},
  number = {1},
  url = {https://skillstore.io/skills/dboeckli-cc-best-practices/audits/1},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "cc-best-practices security audit report (audit version 1)"
version: "unspecified"
type: report
authors:
  - name: "dboeckli"
date-released: "2026-09-26"
url: "https://skillstore.io/skills/dboeckli-cc-best-practices/audits/1"
identifiers:
  - type: other
    value: "skillstore:dboeckli-cc-best-practices:audit:1"
    description: "Skillstore immutable audit report identifier"

Skillstore 评分

评分依据证据置信度: 中

架构

可维护性

内容

社区

规范符合性

你能构建什么

Plan a Complex Change

Structure exploration, planning, implementation, and verification before editing several related files.

Standardize Team Instructions

Create a concise CLAUDE.md that records project-specific commands, architecture, testing rules, and repository conventions.

Scale Repetitive Analysis

Choose subagents, non-interactive commands, or parallel worktrees while preserving focused review and verification.

试试这些提示

Improve a Task Prompt

Rewrite my request with explicit scope, relevant files, constraints, expected behavior, and one runnable verification step.

Prepare an Implementation Plan

Explore the affected code without editing. Then produce a sequenced plan with dependencies, risks, files, and verification commands.

Recover From Repeated Errors

Identify missing context behind repeated failures. Draft a fresh prompt that states the constraint, example, scope, and acceptance test.

Design a Parallel Workflow

Design separate writer and reviewer sessions using worktrees. Define ownership, allowed tools, handoff evidence, and final integration checks.

最佳实践

Provide a runnable check and request evidence from its output.
Separate exploration, planning, implementation, and independent review for complex work.
Keep persistent instructions short, project-specific, and necessary.

避免

Do not combine unrelated tasks in one crowded session.
Do not continue correcting the same failure without improving the original context.
Do not grant broad automation permissions without reviewing data exposure and persistent changes.