{
  "markdown": "# Claude Code Toolkit\n\n> My personal Claude Code commands and agents after 6 months of daily use. Steal freely.\n\n## Plugin Installation (Claude Plugin Directory)\n\nOnce listed in the Claude plugin directory:\n```bash\nclaude plugin install claude-code-toolkit@claude-community\n```\nCommands will be available as `/claude-code-toolkit:build`, `/claude-code-toolkit:plan`, etc.\n\n## What's in here\n\n10 slash commands and 3 agent definitions I actually use. Not a framework — just the stuff that survived.\n\n### Commands that matter most\n\n**`/build`** — Spawns planner → coder → reviewer pipeline. Give it a task, come back to working code.\n\n**`/dispatch`** — Breaks a large task into parallel subtasks and runs them. Works for refactors, audits, migrations.\n\n**`/security-audit`** — Full security pass: hardcoded secrets, SQL injection, auth gaps, dependency vulns. Produces a prioritized findings report.\n\n### All commands\n\n| Command | What it does |\n|---------|-------------|\n| `/build` | Multi-agent build pipeline (planner → coder → reviewer) |\n| `/dispatch` | Parallel task execution |\n| `/security-audit` | Security scan with prioritized findings |\n| `/handoff` | Generate a handoff doc from current session state |\n| `/fresh` | Clear context, start clean without losing work |\n| `/init` | Initialize a new project with standard structure |\n| `/plan` | Generate a phased implementation plan |\n| `/railway-preflight` | Pre-deploy checklist for Railway |\n| `/verify` | Run verification pass on completed work |\n| `/wait-for` | Poll until a condition is met |\n\n### Agents\n\n| Agent | Role |\n|-------|------|\n| `planner.md` | Breaks tasks into phases, writes PLAN.md |\n| `coder.md` | Implements phases, commits after each |\n| `reviewer.md` | Reviews output, flags issues, suggests improvements |\n\n## Install\n\n```bash\n# Clone into your Claude config directory\ngit clone https://github.com/rjspence3/claude-code-toolkit ~/.claude/toolkit\n\n# Copy commands and agents\ncp -r ~/.claude/toolkit/.claude/commands/* ~/.claude/commands/\ncp -r ~/.claude/toolkit/.claude/agents/* ~/.claude/agents/\n```\n\nThat's it. Commands are available immediately in Claude Code as `/command-name`.\n\n## Usage\n\n```bash\n# Build something\n/build implement a rate limiter for the API\n\n# Run a security audit\n/security-audit\n\n# Parallel refactor\n/dispatch refactor all API routes to use the new error handler\n```\n\n## Related Plugins\n\nThese tools were split out into their own standalone plugins:\n\n| Plugin | What it does |\n|--------|-------------|\n| [claude-site-audit](https://github.com/rjspence3/claude-site-audit) | Accessibility + SEO + GEO audit with letter grade and PDF report |\n| [claude-beta-tester](https://github.com/rjspence3/claude-beta-tester) | Personality-driven QA via AI agents (speedrunners, chaos gremlins, executives) |\n\n## What I've learned\n\nAfter 6 months: the commands that actually stick are the ones that handle the annoying parts of a workflow, not the parts that are interesting. Nobody needs a command that makes coding more exciting. You need commands that handle the handoff, the cleanup, the review pass — the parts you'd otherwise skip.\n\n## Background\n\nI wrote about the experiment that shaped these in [this post](https://substack.com/@rjspence3). The prompt format doesn't matter as much as you think. The workflow does.\n\n---\n\nBuilt by [Rob Spence](https://nomouthlabs.com). MIT License.\n",
  "bytes": 3385,
  "sha": "7d277d9264511dda9e09a342d2b4670ad2703d4c9339eb2b545df711152167fc",
  "repo_slug": "rjspence3/claude-code-toolkit",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_rjspence3_claude_code_toolkit_claude_cod_1e8263a6/readme"
}