claude-code-agents

> Production-ready subagents for Claude Code CLI — 8 specialized agents covering security, code quality, testing, SEO, documentation, infrastructure, databases, and API design.

Each agent runs as an independent subagent with its own context window, system prompt, tool permissions, MCP server access, and domain expertise. They can be invoked individually with slash commands, chained into pipelines, or run as a full audit suite.

---

Claude Code supports custom agents — markdown files placed in .claude/agents/ that define specialized personas with their own tools, permissions, and knowledge. When you invoke a subagent:

• It runs in a separate context window from your main conversation
• It has access to only the tools and MCP servers you specify in its frontmatter
• It loads domain-specific knowledge via skills preloading
• It follows a structured methodology defined in its system prompt
• It can enforce safety constraints through hook scripts

This project provides 8 battle-tested agents that cover the most common engineering needs.

| Agent | Tag | Domain | Command | Key Capabilities | |-------|-----|--------|---------|-----------------| | CyberSentinel | SEC | Security | /sec | OWASP Top 10 audit, CVE scanning, penetration testing, secret detection, dependency analysis | | CodeCraft | BPR | Code Quality | /bpr | Architecture review, code smells, SOLID/DRY/KISS, framework-specific best practices | | TestForge | TST | Testing & QA | /tst | Unit/integration/e2e generation, coverage analysis, security regression tests | | GrowthForge | SEO | SEO & Performance | /seo | Technical SEO, Core Web Vitals, structured data, accessibility, conversion optimization | | DocMaster | DOC | Documentation | /doc | README, API docs, changelogs, legal documents, architect