lukehungngo/claude-mas-template
Production-grade multi-agent system template for Claude Code: 7 agents, 13 skills, TDD workflows, and battle-tested engineering practices
ARCHETYPE
Fort Knox
Deny lists, pre-tool hooks, sandboxing. Nothing touches prod without a signature.
GUARDRAILS
4/5
Blocks destructive commands · Protects secrets · Pre-tool screening hook · No YOLO mode · details
Copy this rig
# review before running: this installs third-party code
$ npx degit lukehungngo/claude-mas-template/.claude ./rig-claude-mas-template # inspect, then merge into .claude/ 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.
$ npx degit lukehungngo/claude-mas-template/skills/ask-questions .claude/skills/ask-questions $ npx degit lukehungngo/claude-mas-template/skills/differential-review .claude/skills/differential-review $ npx degit lukehungngo/claude-mas-template/skills/finishing-branch .claude/skills/finishing-branch $ npx degit lukehungngo/claude-mas-template/skills/obsidian .claude/skills/obsidian $ npx degit lukehungngo/claude-mas-template/skills/property-based-testing .claude/skills/property-based-testing $ npx degit lukehungngo/claude-mas-template/skills/reliability-review .claude/skills/reliability-review $ npx degit lukehungngo/claude-mas-template/skills/se-principles .claude/skills/se-principles $ npx degit lukehungngo/claude-mas-template/skills/subagent-driven-development .claude/skills/subagent-driven-development $ npx degit lukehungngo/claude-mas-template/skills/verification .claude/skills/verification
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"deny": [
"Read:.env*",
"Bash:sudo*",
"Bash:rm -rf /*",
"Bash:git push --force*"
]
},
"hooks": {
"PreToolUse": [
{
"matcher": "Agent",
"hooks": [
{
"type": "command",
"command": "$CLAUDE_PROJECT_DIR/.claude/hooks/validate-dispatch.sh"
}
]
},
{
"matcher": "Skill",
"hooks": [
{
"type": "command",
"command": "$CLAUDE_PROJECT_DIR/.claude/hooks/validate-skill.sh"
}
]
},
{
"matcher": "Edit|Write|Bash",
"hooks": [
{
"type": "command",
"command": "$CLAUDE_PROJECT_DIR/hooks/suggest-compact.sh"
}
]
}
]
}
} Skills (9)
ask-questionsdifferential-reviewfinishing-branchobsidianproperty-based-testingreliability-reviewse-principlessubagent-driven-developmentverification
Hooks (6)
| event | matcher | runs |
|---|---|---|
| PreToolUse | Agent | $CLAUDE_PROJECT_DIR/.claude/hooks/validate-dispatch.sh |
| PreToolUse | Skill | $CLAUDE_PROJECT_DIR/.claude/hooks/validate-skill.sh |
| PreToolUse | Edit|Write|Bash | $CLAUDE_PROJECT_DIR/hooks/suggest-compact.sh |
| PostToolUse | Edit|Write | $CLAUDE_PROJECT_DIR/.claude/hooks/lint.sh |
| Stop | * | $CLAUDE_PROJECT_DIR/.claude/hooks/pre-stop-gate.sh |
| Stop | * | $CLAUDE_PROJECT_DIR/hooks/validate-pipeline.sh |
Slash commands (8)
/audit/bootstrap/brainstorm/bug-fix/dev-loop/loop/reflect/release
Plugins (1)
superpowers@claude-plugins-official
Permissions
deny (4)
Read:.env*
Bash:sudo*
Bash:rm -rf /*
Bash:git push --force*
ask (0)
—
allow (18)
Read:*
Write:*
Edit:*
Bash:*
Glob:*
Grep:*
Agent:*
WebSearch
WebFetch:*
Skill:*
TaskCreate
TaskUpdate
TaskGet
TaskList
TaskOutput
TaskStop
NotebookEdit:*
SendMessage:*
Similar rigs
bitbonsai/mcpvault
A lightweight Model Context Protocol (MCP) server for safe Obsidian vault access
Pragmatist 2.6k tok ·
4esv/dotfiles
Personal dotfiles for macOS terminal development. Managed with GNU Stow.
YOLO Cowboy 1.0k tok ·
Aurite-ai/agent-verifier
Agent Verifier is a coding agent skill that verifies code against organizational policies, code quality patterns, security requirements, and framework best practices — before code ships. Works with Claude Code, Cursor, Windsurf, and 30+ age
Pragmatist 313 tok ·
EdgeCaseLabs/claude-code-config
—
Pragmatist 2.2k tok ·