Shehabov/ai-dev-crew
A plug-and-play product team for Claude Code. Sixteen agents plan, design, build, review, test and release your product, with an independent gate at every handoff.
ARCHETYPE
Fort Knox
Deny lists, pre-tool hooks, sandboxing. Nothing touches prod without a signature.
GUARDRAILS
4/5
Blocks destructive commands · Protects secrets · No YOLO mode · Sandbox or ask-first rules · details
Copy this rig
# review before running: this installs third-party code $ claude mcp add playwright -- npx @playwright/mcp@0.0.82 --no-webmcp $ claude mcp add --transport http supabase 'https://mcp.supabase.com/mcp?project_ref=<project-ref>&features=docs%2Caccount%2Cdatabase%2Cdebugging%2Cdevelopment%2Cfunctions%2Cbranching' $ npx degit Shehabov/ai-dev-crew/.claude ./rig-ai-dev-crew # 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.
$ claude mcp add playwright -- npx @playwright/mcp@0.0.82 --no-webmcp $ claude mcp add --transport http supabase 'https://mcp.supabase.com/mcp?project_ref=<project-ref>&features=docs%2Caccount%2Cdatabase%2Cdebugging%2Cdevelopment%2Cfunctions%2Cbranching'
$ npx degit Shehabov/ai-dev-crew/.claude/skills/stack-nextjs-supabase .claude/skills/stack-nextjs-supabase $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-architecture .claude/skills/team-architecture $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-brand-guard .claude/skills/team-brand-guard $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-bug-register .claude/skills/team-bug-register $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-clean-code .claude/skills/team-clean-code $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-code-analysis .claude/skills/team-code-analysis $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-code-review .claude/skills/team-code-review $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-copy .claude/skills/team-copy $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-design-system .claude/skills/team-design-system $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-orchestration .claude/skills/team-orchestration $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-protocol .claude/skills/team-protocol $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-release .claude/skills/team-release $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-security .claude/skills/team-security $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-test-protocol .claude/skills/team-test-protocol $ npx degit Shehabov/ai-dev-crew/.claude/skills/team-ux-audit .claude/skills/team-ux-audit
$ curl -fsSL --create-dirs -o .claude/agents/backend-engineer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/backend-engineer.md $ curl -fsSL --create-dirs -o .claude/agents/bug-historian.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/bug-historian.md $ curl -fsSL --create-dirs -o .claude/agents/code-analyst.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/code-analyst.md $ curl -fsSL --create-dirs -o .claude/agents/code-steward.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/code-steward.md $ curl -fsSL --create-dirs -o .claude/agents/engineering-lead.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/engineering-lead.md $ curl -fsSL --create-dirs -o .claude/agents/frontend-engineer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/frontend-engineer.md $ curl -fsSL --create-dirs -o .claude/agents/orchestrator.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/orchestrator.md $ curl -fsSL --create-dirs -o .claude/agents/peer-reviewer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/peer-reviewer.md $ curl -fsSL --create-dirs -o .claude/agents/qc-engineer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/qc-engineer.md $ curl -fsSL --create-dirs -o .claude/agents/qc-lead.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/qc-lead.md $ curl -fsSL --create-dirs -o .claude/agents/release-engineer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/release-engineer.md $ curl -fsSL --create-dirs -o .claude/agents/security-analyst.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/security-analyst.md $ curl -fsSL --create-dirs -o .claude/agents/tech-architect.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/tech-architect.md $ curl -fsSL --create-dirs -o .claude/agents/ux-auditor.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/ux-auditor.md $ curl -fsSL --create-dirs -o .claude/agents/ux-designer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/ux-designer.md $ curl -fsSL --create-dirs -o .claude/agents/ux-writer.md https://raw.githubusercontent.com/Shehabov/ai-dev-crew/main/.claude/agents/ux-writer.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"deny": [
"Bash(git push --force:*)",
"Bash(git push -f:*)",
"Bash(git reset --hard:*)",
"Bash(git clean -fd:*)",
"Bash(rm -rf:*)",
"Read(./.env)",
"Read(./.env.local)",
"Read(./.env.*.local)",
"Read(./**/.env)",
"Read(./**/.env.local)",
"Read(./**/.env.*.local)",
"Read(./**/secrets/**)",
"mcp__playwright__browser_run_code_unsafe",
"Read(~/.claude/.credentials.json)"
],
"ask": [
"Bash(git commit:*)",
"Bash(git push:*)",
"Bash(git tag:*)",
"Bash(gh pr:*)",
"Bash(gh release:*)",
"Bash(npm publish:*)",
"mcp__supabase__create_branch",
"mcp__supabase__delete_branch",
"mcp__supabase__merge_branch",
"mcp__supabase__reset_branch",
"mcp__supabase__rebase_branch",
"mcp__supabase__update_storage_config",
"mcp__supabase__pause_project",
"mcp__supabase__restore_project",
"mcp__supabase__create_project",
"mcp__supabase__get_cost",
"mcp__supabase__confirm_cost",
"mcp__playwright__browser_file_upload",
"mcp__playwright__browser_drop",
"Edit(.devteam/bin/**)"
]
}
} MCP servers (2)
| server | source | est. tokens |
|---|---|---|
| Playwright MCP · "playwright" | npm | 7.5k |
| Supabase | remote · remote | 7.0k |
Skills (15)
stack-nextjs-supabaseteam-architectureteam-brand-guardteam-bug-registerteam-clean-codeteam-code-analysisteam-code-reviewteam-copyteam-design-systemteam-orchestrationteam-protocolteam-releaseteam-securityteam-test-protocolteam-ux-audit
Subagents (16)
| backend-engineer model: inherit | Use this agent when back-end work has to be built against a tech-architect task brief, on whatever stack PROJECT.md names. That covers schema and hand-authored migrations, access policies and grants, |
| bug-historian model: inherit | Use this agent at the start of every run, before any other agent plans, to brief the team on defects and agent mistakes already recorded in BUGS.md against the surfaces this change touches. Use it aga |
| code-analyst model: inherit | Use this agent when a change needs a mechanical, line-by-line defect and structural-rot scan before it reaches the engineering lead, normally right after backend-engineer or frontend-engineer report a |
| code-steward model: inherit | Use this agent as one of the four independent review gates at stage 6, in parallel with peer-reviewer, code-analyst and security-analyst, on every change that touches code. It enforces the clean code |
| engineering-lead model: inherit | Use this agent when a change has cleared all four independent reviews (peer-reviewer, code-analyst, code-steward, security-analyst) and the bug-historian regression guard, and needs the engineering ga |
| frontend-engineer model: inherit | Use this agent when interface code has to be written or changed against an existing tech-architect frontend brief and an audited ux-designer spec, on whatever stack PROJECT.md names. That covers scree |
| orchestrator model: inherit | Use this agent when any change to the product needs to run end to end across the team, from a brief by the Product Lead through design, build, review, test and release. It interviews the Product Lead |
| peer-reviewer model: inherit | Use this agent when backend-engineer or frontend-engineer has finished implementing against a task brief and the change needs a senior engineering judgement pass before it reaches the integration gate |
| qc-engineer model: inherit | Use this agent when any change has been through the engineering-lead's integration gate and needs to be tested before it can ship, or when a defect report needs reproduction and triage. It tests the A |
| qc-lead model: inherit | Use this agent when the qc-engineer has finished a test pass and produced an evidence set, when a run needs its final independent quality gate before anything reaches the Product Lead, or when anyone |
| release-engineer model: inherit | Use this agent when a change has cleared the qc-lead's quality gate and needs to be released, committed, tagged, verified against its target, or rolled back. It is the only role permitted to push to a |
| security-analyst model: inherit | Use this agent on every change, every build and every commit, without exception, and again before any release. It is the data and code security gate, covering exposed keys and credentials in the worki |
| tech-architect model: inherit | Use this agent when any change to the product is proposed, because the architecture is re-examined for every change, including changes that add no feature. It runs at stage 1, straight after the bug-h |
| ux-auditor model: inherit | Use this agent when a ux-designer handoff needs independent verification before the design gate, when shipped UI needs an adversarial audit against the brand spec, the accessibility standard in PROJEC |
| ux-designer model: inherit | Use this agent when a surface needs to be designed or redesigned before anyone writes code, when the tech-architect has issued a frontend task brief that implies a new screen, state, flow or component |
| ux-writer model: inherit | Use this agent when any user-visible string is being created, changed, localised or reviewed, in any locale named in PROJECT.md. Trigger it when the tech-architect issues a task brief that touches a s |
Permissions
deny (14)
Bash(git push --force:*)
Bash(git push -f:*)
Bash(git reset --hard:*)
Bash(git clean -fd:*)
Bash(rm -rf:*)
Read(./.env)
Read(./.env.local)
Read(./.env.*.local)
Read(./**/.env)
Read(./**/.env.local)
Read(./**/.env.*.local)
Read(./**/secrets/**)
mcp__playwright__browser_run_code_unsafe
Read(~/.claude/.credentials.json)
ask (20)
Bash(git commit:*)
Bash(git push:*)
Bash(git tag:*)
Bash(gh pr:*)
Bash(gh release:*)
Bash(npm publish:*)
mcp__supabase__create_branch
mcp__supabase__delete_branch
mcp__supabase__merge_branch
mcp__supabase__reset_branch
mcp__supabase__rebase_branch
mcp__supabase__update_storage_config
mcp__supabase__pause_project
mcp__supabase__restore_project
mcp__supabase__create_project
mcp__supabase__get_cost
mcp__supabase__confirm_cost
mcp__playwright__browser_file_upload
mcp__playwright__browser_drop
Edit(.devteam/bin/**)
allow (68)
Read
Glob
Grep
Write(.devteam/**)
Edit(.devteam/**)
Bash(git status:*)
Bash(git diff:*)
Bash(git log:*)
Bash(git show:*)
Bash(git branch:*)
Bash(git rev-parse:*)
Bash(git ls-files:*)
Bash(git grep:*)
Bash(git fetch:*)
Bash(git merge-base:*)
Bash(git add:*)
Bash(git stash:*)
Bash(git --version)
Bash(node --version)
Bash(npm --version)
Bash(npx --version)
Bash(claude mcp list:*)
Bash(npm install:*)
Bash(npm ci:*)
Bash(npm run:*)
Bash(npm test:*)
Bash(npm ls:*)
Bash(npm audit:*)
Bash(npx tsc:*)
Bash(npx next:*)
Bash(npx vitest:*)
Bash(npx playwright:*)
Bash(npx create-next-app:*)
Bash(node .devteam/bin/*)
Bash(node .devteam/bin/sync-gates.mjs:*)
Bash(node .devteam/bin/utilisation-check.mjs:*)
Bash(node .claude/skills/*/scripts/*)
Bash(node scripts/check.mjs)
Bash(ls:*)
Bash(cat:*)
Bash(head:*)
Bash(tail:*)
Bash(wc:*)
Bash(find:*)
Bash(grep:*)
Bash(sort:*)
Bash(date:*)
Bash(mkdir:*)
mcp__supabase__search_docs
mcp__supabase__list_tables
mcp__supabase__list_extensions
mcp__supabase__list_migrations
mcp__supabase__apply_migration
mcp__supabase__execute_sql
mcp__supabase__get_logs
mcp__supabase__get_advisors
mcp__supabase__get_project_url
mcp__supabase__get_publishable_keys
mcp__supabase__get_anon_key
mcp__supabase__generate_typescript_types
Similar rigs
Jailsom-Nogueira/nextjs-bootstrap-template
Agent-ready Next.js 16 + Supabase starter with auth, i18n, analytics, email, QA loops, and multi-agent coding docs.
Pragmatist 19.1k tok ·
maltehedderich/master-class-agents
A curated collection of AI agent templates, built on the architectural principles and design patterns taught by the most influential voices in the field.
Orchestrator 2.6k tok ·
mcp-tool-shop-org/role-os
Role OS — a repo-native operating layer that staffs coding-agent work through 61 role contracts, packs, missions, recipe cards, and a critic jury. Claude Code adapter ships; the contracts are markdown any harness can consume.
Orchestrator 364 tok ·
evgenii-studitskikh/Claude-Code-SaaS-Studio
Turn Claude Code into a complete SaaS development studio with AI agents for product management, UX, Next.js, Supabase, Stripe, DevOps, and launch workflows.
Orchestrator 1.8k tok ·