royfrenk/claude-config
Multi-agent workflow system for Claude Code
ARCHETYPE
YOLO Cowboy
Permissions? Never heard of 'em. Ships at the speed of `--dangerously-skip-permissions`.
Copy this rig
# review before running: this installs third-party code $ npx degit royfrenk/claude-config/agents ./rig-claude-config/agents $ npx degit royfrenk/claude-config/commands ./rig-claude-config/commands $ npx degit royfrenk/claude-config/hooks ./rig-claude-config/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ curl -fsSL --create-dirs -o .claude/agents/design-planner.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/design-planner.md $ curl -fsSL --create-dirs -o .claude/agents/design-reviewer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/design-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/developer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/developer.md $ curl -fsSL --create-dirs -o .claude/agents/em.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/em.md $ curl -fsSL --create-dirs -o .claude/agents/eval-writer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/eval-writer.md $ curl -fsSL --create-dirs -o .claude/agents/explorer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/explorer.md $ curl -fsSL --create-dirs -o .claude/agents/external-model-delegate.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/external-model-delegate.md $ curl -fsSL --create-dirs -o .claude/agents/linear-sync.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/linear-sync.md $ curl -fsSL --create-dirs -o .claude/agents/mobile-verifier.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/mobile-verifier.md $ curl -fsSL --create-dirs -o .claude/agents/plan-writer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/plan-writer.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/scraping-agent.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/scraping-agent.md $ curl -fsSL --create-dirs -o .claude/agents/security-reviewer.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/security-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/sre.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/sre.md $ curl -fsSL --create-dirs -o .claude/agents/visual-verifier.md https://raw.githubusercontent.com/royfrenk/claude-config/main/agents/visual-verifier.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"hooks": {
"PreToolUse": [
{
"matcher": "*",
"hooks": [
{
"type": "command",
"command": "START_FILE=\"/tmp/claude-session-start\"; if [ ! -f \"$START_FILE\" ]; then date +%s > \"$START_FILE\"; fi"
}
]
}
]
}
} Subagents (15)
| design-planner model: sonnet | Creates design specifications BEFORE Explorer for ANY UI/UX work (new or existing). Validates all links and external references with User. |
| design-reviewer model: sonnet | Reviews UI/UX implementations against design standards using code review + screenshot verification. Mandatory for all UI work. |
| developer model: sonnet | Code implementation and deployment. Use proactively for writing code, fixing bugs, running tests, and deploying to staging. Executes tasks assigned by eng-manager. |
| em model: none (runs in main conversation) | Engineering Manager — sprint orchestration, task coordination, roadmap management. Orchestrates Explorer, Plan-Writer, Developer, Reviewer, and SRE agents. |
| eval-writer model: sonnet | Writes quality evaluation criteria for features with subjective/ranking/performance requirements. Human-verified. |
| explorer model: sonnet | Codebase analysis and technical specification creation |
| external-model-delegate model: sonnet | Delegates bug fixes to external AI models (Gemini, Codex). Generates context, calls external model API, implements the suggested fix. Spawned by /iterate. |
| linear-sync model: haiku | Handles Linear synchronization with graceful failure handling and timeouts |
| mobile-verifier model: sonnet | Captures screenshots and performs interactive functional + UX/behavioral verification on iOS simulator (and physical device via maestro-runner) using Maestro. The mobile counterpart to visual-verifier |
| plan-writer model: sonnet | Implementation planning based on Explorer's findings |
| reviewer model: sonnet | Code review specialist. Use proactively to review code changes, check for quality issues, security problems, and approve staging deployments. Reviews Developer submissions. |
| scraping-agent | Web scraping specialist. Use when the user wants to scrape a website, add a new data source, check scraping status, debug blocked requests, queue new jobs, or manage the worker library. Knows how to b |
| security-reviewer model: sonnet | Security review specialist. Reviews code diffs for security issues during sprints (PR mode) and runs full codebase audits periodically (audit mode). Invoked by EM after Code Reviewer approves. |
| sre model: claude-sonnet-4-6 | Site Reliability Engineer — deployment monitoring, health checks, failure analysis. Uses Managed Agents (Anthropic-hosted). Per-project opt-in via .sre/config.yaml. |
| visual-verifier model: sonnet | Captures screenshots and performs interactive visual verification using Playwright. Supports static capture (breakpoints) and interactive flows (navigate, click, fill, verify). Invoked by EM for desig |
Hooks (11)
| event | matcher | runs |
|---|---|---|
| PreToolUse | * | START_FILE="/tmp/claude-session-start"; if [ ! -f "$START_FILE" ]; then date +%s > "$START_FILE"; fi |
| PostToolUse | tool == 'Edit' && tool_input.file_path matches '\.(ts|tsx|js|jsx)$' | prettier --write "$CLAUDE_FILE_PATH" 2>/dev/null || true |
| PostToolUse | tool == 'Edit' && tool_input.file_path matches '\.(ts|tsx|js|jsx)$' | grep -n 'console\.log' "$CLAUDE_FILE_PATH" 2>/dev/null && echo '[Warning] console.log found - remove before committing' >&2 || true |
| PostToolUse | tool == 'Edit' && tool_input.file_path matches '\.(ts|tsx)$' | if [ -f "$(dirname "$CLAUDE_FILE_PATH")/tsconfig.json" ] || [ -f "$(git rev-parse --show-toplevel 2>/dev/null)/tsconfig.json" ]; then cd "$(git rev-parse --show-toplevel 2>/dev/null)" && npx tsc --noEmit 2>&1 | grep -E 'error TS' | head -5 |
| PostToolUse | tool == 'Edit' && tool_input.file_path matches '\.(ts|tsx|js|jsx|py)$' | grep -nE '(sk-[a-zA-Z0-9]{20,}|api_key\\s*=\\s*["\x27][^"\x27]+["\x27]|password\\s*=\\s*["\x27][^"\x27]+["\x27])' "$CLAUDE_FILE_PATH" 2>/dev/null && echo '[SECURITY] Possible hardcoded secret detected!' >&2 || true |
| PostToolUse | Edit|Write | ~/.claude/hooks/file-size-guard.sh |
| PostToolUse | Write | ~/.claude/hooks/naming-convention-guard.sh |
| PostToolUse | Edit|Write | ~/.claude/hooks/checkpoint-reminder-edits.sh |
| PostToolUse | tool matches '^mcp__linear__' | bash ~/.claude/scripts/cleanup-linear-mcp.sh 2>/dev/null || true |
| Stop | * | afplay /System/Library/Sounds/Ping.aiff & |
| PreCompact | * | ~/.claude/hooks/checkpoint-reminder-compact.sh |
Slash commands (23)
/app-store-opportunity-research/argue/audit/change-process/checkpoint/context/create-bug/create-issue/deep-research/design/iterate/learning-opportunity/new-project/post-mortem/reset-linear/review-prd/security-audit/sprint/sync-linear/sync-roadmap/transcribe/ux-audit/write-prd
Plugins (6)
frontend-design@claude-plugins-officialplayground@claude-plugins-officialcode-simplifier@claude-plugins-officialvercel@claude-plugins-officialsupabase@claude-plugins-officialchrome-devtools-mcp@claude-plugins-official
Permissions
deny (0)
—
ask (0)
—
allow (21)
Bash
Read
Write
Edit
Glob
Grep
WebFetch
WebSearch
mcp__*
mcp__linear__get_issue
mcp__linear__update_issue
mcp__linear__list_issues
mcp__linear__create_comment
mcp__google-docs__searchDriveFiles
Bash(touch ~/.claude/agents/em-write-test.txt)
Bash(rm ~/.claude/agents/em-write-test.txt)
Bash(cp -R ~/.claude/agents ~/Documents/repos/claude-config/)
Bash(cp -R ~/.claude/commands ~/Documents/repos/claude-config/)
Bash(cp -R ~/.claude/rules ~/Documents/repos/claude-config/)
Bash(cp -R ~/.claude/guides ~/Documents/repos/claude-config/)
Bash(cp -R ~/.claude/scripts ~/Documents/repos/claude-config/)
Similar rigs
jaqubowsky/fleet
Runs Claude Code and pi agents in parallel on your Mac, one Docker sandbox per issue. A host agent hands out the work, watches every sandbox and brings back tested, reviewed pull requests.
Minimalist 341 tok ·
pahlevikun/levistack
Agent skills for people tired of AI slop. 35 skills, 11 subagents, 41 rules and 5 hooks for Claude Code, Cursor and Codex. Verifies before it says done.
Orchestrator 3.1k tok ·
joshido/ai-template
A project-agnostic four-agent AI development workflow template for Claude Code, Gemini CLI, and GitHub Copilot.
Pragmatist 1.3k tok ·
itshig/AGENTS-template
Drop-in AGENTS.md template for Claude Code, Cursor, Codex, and other AI agents — with .agents/ for rules, personas, and workflows.
Orchestrator 3.0k tok ·