{"id":50478698,"url":"https://github.com/deepakb/antigravity-studio","last_synced_at":"2026-06-01T15:32:12.333Z","repository":{"id":345065529,"uuid":"1184252864","full_name":"deepakb/antigravity-studio","owner":"deepakb","description":"A high-performance CLI and toolkit for deploying specialized AI agents, expert skills, and automated quality gates in TypeScript projects.","archived":false,"fork":false,"pushed_at":"2026-04-14T12:47:42.000Z","size":890,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"master","last_synced_at":"2026-04-14T14:30:50.533Z","etag":null,"topics":["ai-agents","ai-engineering","automation","cli-tool","developer-tools","enterprise-ai","nextjs","quality-gates","react","typescript"],"latest_commit_sha":null,"homepage":"","language":"TypeScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/deepakb.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-03-17T12:07:30.000Z","updated_at":"2026-04-14T12:47:48.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/deepakb/antigravity-studio","commit_stats":null,"previous_names":["deepakb/antigravity-studio"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/deepakb/antigravity-studio","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepakb%2Fantigravity-studio","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepakb%2Fantigravity-studio/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepakb%2Fantigravity-studio/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepakb%2Fantigravity-studio/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/deepakb","download_url":"https://codeload.github.com/deepakb/antigravity-studio/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/deepakb%2Fantigravity-studio/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33782314,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-06-01T02:00:06.963Z","response_time":115,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-agents","ai-engineering","automation","cli-tool","developer-tools","enterprise-ai","nextjs","quality-gates","react","typescript"],"created_at":"2026-06-01T15:32:11.419Z","updated_at":"2026-06-01T15:32:12.326Z","avatar_url":"https://github.com/deepakb.png","language":"TypeScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# ⚡ Nexus Studio\n\n\u003e **The AI Dev OS for Every Team** — One `npx` command installs 29 specialist AI agents, 51 expert skill libraries, and 12 automated quality gates into any project. Works with Cursor, Windsurf, Copilot, and Claude Code — consistent AI behaviour across every developer on your team.\n\n[![npm version](https://img.shields.io/npm/v/@nexus/studio.svg)](https://www.npmjs.com/package/@nexus/studio)\n[![Node.js](https://img.shields.io/badge/node-%3E%3D18.0.0-brightgreen)](https://nodejs.org)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)\n[![TypeScript](https://img.shields.io/badge/TypeScript-5.7-blue)](https://www.typescriptlang.org)\n[![Tests](https://img.shields.io/badge/tests-43%20passed-brightgreen)](#-testing)\n[![PRs Welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](#-contributing)\n\n---\n\n## 🗺️ Quick Navigation\n\n| I want to… | Go here |\n|---|---|\n| Bootstrap a new project in 60 seconds | [🚀 Quick Start](#-quick-start) |\n| Understand how the system works | [🏛️ Architecture](#️-architecture) |\n| See all 24 CLI commands | [💻 CLI Reference](#-cli-reference) |\n| Enforce standards across a team | [🏢 Enterprise Configuration](#-enterprise-configuration) |\n| Audit an existing project | [`studio doctor`](#studio-doctor) · [`studio validate`](#studio-validate) |\n| Browse agents, skills \u0026 profiles | [🤖 Agents](#-specialist-agents-29) · [🧠 Skills](#-expert-skills-51) · [📦 Profiles](#-supported-stacks--profiles) |\n| Contribute a skill or agent | [🤝 Contributing](#-contributing) |\n\n---\n\n## What Is Nexus Studio?\n\nNexus Studio is a polyglot enterprise CLI that installs a **local AI agent system** (`.agent/`) into any project. It gives every developer on your team the same specialist AI agents, skill libraries, and automated quality gates — regardless of IDE, framework, or language.\n\n**The problem it solves:** Every developer has a different AI setup — one uses Cursor with custom rules, another Copilot with no context, another Claude with a half-written system prompt. The result: inconsistent AI behaviour across the team. Nexus Studio fixes this with a **single canonical `.agent/` source of truth** that every IDE generates its config from.\n\n```\n                         ┌─────────────────────────────────┐\n                         │         Nexus Studio CLI         │\n                         │    npx @nexus/studio init        │\n                         └──────────────┬──────────────────┘\n                                        │  detects + installs\n                         ┌──────────────▼──────────────────┐\n                         │          .agent/  (local)        │\n                         │  agents/ · skills/ · workflows/  │\n                         │  scripts/ · mcp/                 │\n                         └──┬─────────┬──────┬──────┬──────┘\n                            │         │      │      │\n                   ┌────────▼─┐ ┌─────▼──┐ ┌▼────┐ ┌▼──────────┐\n                   │  Cursor  │ │Windsurf│ │Copil│ │Claude Code│\n                   │  .mdc    │ │.rules  │ │ot   │ │CLAUDE.md  │\n                   └──────────┘ └────────┘ └─────┘ └───────────┘\n```\n\n---\n\n## 📊 Registry at a Glance\n\n| | Count |\n|---|---|\n| 🤖 Specialist Agents | **29** across 8 domains |\n| 🧠 Expert Skills | **51** across 13 categories |\n| 🗂️ Architectural Profiles | **15** + Custom |\n| ✅ Quality Gates | **12** (3 enforcement tiers) |\n| ⚡ Slash Commands | **15** |\n| 🌐 Polyglot Stacks | **5** (Node · Python · Java · .NET · Flutter) |\n| 🖥️ IDE Integrations | **4** (Cursor · Windsurf · Copilot · Claude) |\n\n---\n\n## 🚀 Quick Start\n\n```bash\n# Run once in any project — no global install needed\nnpx @nexus/studio init\n\n# Or install globally\nnpm install -g @nexus/studio\nstudio init\n```\n\n**What you see during `studio init`:**\n\n```\n  ⚡ Nexus Studio — The AI Dev OS for Every Team\n\n  ✔ Detected: Next.js 15 · TypeScript · Tailwind CSS · Prisma\n  ✔ Suggested profile: nextjs-fullstack\n\n  ? Select your IDE integrations: (Space to toggle)\n    ◉ GitHub Copilot\n    ◉ Cursor\n    ○ Windsurf\n    ○ Claude Code\n\n  Installing 29 agents · 51 skills · 15 workflows...\n  ████████████████████████████ 100%\n\n  ✔ .agent/         installed (147 files)\n  ✔ .github/copilot-instructions.md generated\n  ✔ .cursor/rules/nexus-system.mdc generated\n  ✔ .git/info/exclude updated (.agent/ is local-only — never committed)\n  ✔ .agstudio.json created\n\n  ✨ Ready! Your AI agents are active. Try /blueprint in your IDE.\n```\n\nThe `init` wizard will:\n1. **Auto-detect** your framework (Next.js, Python/FastAPI, Java/Spring, etc.)\n2. **Suggest a profile** with the right agent + skill set\n3. **Install `.agent/`** — your local AI knowledge base\n4. **Generate IDE config** — Cursor rules, Copilot instructions, Claude agents, etc.\n5. **Wire Git exclusions** — `.agent/` stays local, never committed\n\n---\n\n## 🏛️ Architecture\n\nNexus Studio is built on a **three-layer decoupled architecture**:\n\n### Layer 1 — The CLI Orchestrator (`studio`)\nA lightweight TypeScript CLI (`@nexus/studio`) that handles detection, installation, drift management, and quality gate execution. It never prescribes implementation details — it only wires the agent system into your project.\n\n### Layer 2 — The Agent System (`.agent/`)\nA project-local directory that is the **single source of truth** for all AI behaviour. It is IDE-agnostic and survives IDE changes. Each IDE generator reads from `.agent/` and produces its own config format.\n\n```\n.agent/\n├── AGENT_SYSTEM.md      # Master system prompt \u0026 operating directives\n├── AGENTS.md            # Agent routing index \u0026 coalition patterns\n├── AGENT_FLOW.md        # 5-stage execution pipeline definition\n├── ARCHITECTURE.md      # Project architecture reference\n├── agents/              # 29 specialist agent personas (.md)\n├── skills/              # 51 expert skill libraries (.md)\n├── workflows/           # 15 slash command procedures (.md)\n├── scripts/             # 12 quality gate shell runners (.sh per stack)\n└── mcp/                 # MCP server configuration\n```\n\n### Layer 3 — The Registry (`registry.json`)\nA version-controlled manifest that defines all agents, skills, profiles, and slash commands. The `sync` and `update` commands use SHA-256 hash drift detection to identify when upstream templates have changed.\n\n---\n\n## � How a Request Flows — 8-Stage Pipeline\n\nWhen you type a prompt in your IDE, Nexus routes it through a structured pipeline defined in `.agent/AGENT_FLOW.md`:\n\n```\n  1. REQUEST ENTRY    ─  Natural language · slash command · domain request\n          │\n  2. SOCRATIC GATE    ─  New feature? → 3 strategic clarifying questions\n          │              Bug fix?     → confirm scope → proceed\n          │\n  3. AGENT SELECTION  ─  Auto-routes via AGENTS.md routing table\n          │              (no @mention needed — intent detection is automatic)\n          │\n  4. AGENT INIT       ─  Loads persona · behavioral constraints · output format\n          │\n  5. SKILL LOADING    ─  Matches request → injects domain SKILL.md into context\n          │              Cross-links related skills (e.g. owasp-top10 ↔ auth-nextauth)\n          │\n  6. TASK EXECUTION   ─  Agent executes with full combined context\n          │              Scripts require explicit user approval before running\n          │\n  7. VALIDATION       ─  Quick (~30s): security · lint/TS · schema · unit tests\n          │              Full  ( ~5m): + Lighthouse · Playwright E2E · bundle · i18n\n          │\n  8. RESULT DELIVERY  ─  Code diff + explanation + next-step suggestions\n```\n\n\u003e **Multi-agent note:** When a request spans multiple domains (e.g. UI + API + DB), agents execute **sequentially** with context hand-off — not in true parallel. The AI processes one domain at a time.\n\n---\n\n## �📦 Supported Stacks \u0026 Profiles\n\nNexus Studio auto-detects your project and maps it to the most appropriate profile:\n\n| Profile ID | Description | Stack Runner |\n|---|---|---|\n| `nextjs-fullstack` ⭐ | Next.js + Backend + DB | Node |\n| `nextjs-frontend` | React/Next.js — UI only | Node |\n| `react-vite` | React + Vite SPA | Node |\n| `angular-enterprise` | Angular with RxJS/NgRx | Node |\n| `vue-nuxt` | Full-stack Vue with Nuxt | Node |\n| `vue-vite` | Client-side Vue SPA | Node |\n| `expo-mobile` | React Native / Expo | Node |\n| `node-api` | Node.js REST/GraphQL API | Node |\n| `monorepo-root` | Turborepo / Nx workspace root | Node |\n| `monorepo-package` | Sub-package within a monorepo | Node |\n| `python-fastapi` | Async Python REST API | Python |\n| `python-django` | Full-stack Django application | Python |\n| `java-spring` | Enterprise Java / Spring Boot | Java |\n| `dotnet-api` | .NET / ASP.NET Core API | .NET |\n| `flutter-mobile` | Cross-platform Flutter app | Flutter |\n| `custom` | Hand-pick individual agents + skills | Any |\n\n---\n\n## 🤖 Specialist Agents (29)\n\nAgents are expert personas that the AI activates automatically based on request context. They are routed via the `AGENTS.md` auto-routing table — no manual `@mention` required.\n\n### 🏛️ Architecture \u0026 Leadership (6)\n`enterprise-architect` · `tech-lead` · `orchestrator` · `product-manager` · `project-planner` · `cloud-architect`\n\n### 🌐 Frontend — Web (4)\n`nextjs-expert` · `react-performance-guru` · `frontend-specialist` · `ui-component-architect`\n\n### 🎨 UX / UI Design (3)\n`ux-designer` · `ui-design-engineer` · `accessibility-auditor`\n\n### ⚙️ Backend \u0026 API (4)\n`backend-specialist` · `api-architect` · `database-engineer` · `data-layer-specialist`\n\n### 🛡️ Security (2)\n`security-engineer` · `penetration-tester`\n\n### 📱 Mobile (3)\n`rn-architect` · `rn-performance-expert` · `mobile-ux-designer`\n\n### ✅ Quality \u0026 Ops (4)\n`qa-engineer` · `devops-engineer` · `debugger` · `seo-specialist`\n\n### 🤖 AI \u0026 Innovation (3)\n`ai-engineer` · `prompt-engineer` · `llm-security-officer`\n\n\u003e **LLM Safety Rule**: Any request touching AI/LLM features **always** auto-activates `@llm-security-officer` as a silent co-reviewer — no exceptions.\n\n---\n\n## 🧠 Expert Skills (51)\n\n\u003cdetails\u003e\n\u003csummary\u003e51 skills across 13 categories — click to expand\u003c/summary\u003e\n\nSkills are domain-specific coding guidelines injected into the agent's context. Each skill has a token budget to keep context lean. Use `studio info skill \u003cid\u003e` to inspect any skill in full detail.\n\n| Category | Skills |\n|---|---|\n| **Architecture** | `clean-architecture` · `solid-principles` · `monorepo-turborepo` |\n| **Frontend \u0026 UI** | `nextjs-app-router` · `react-patterns` · `shadcn-radix-ui` · `tailwind-design-system` · `angular-patterns` · `rxjs-patterns` · `vue-patterns` · `form-handling` · `framer-motion` · `dark-mode-theming` · `i18n-localization` · `state-management` · `accessibility-wcag` |\n| **Backend \u0026 API** | `api-design-restful` · `prisma-orm` · `realtime-patterns` · `node-express-testing` · `python-fastapi-patterns` · `python-django-patterns` · `java-spring-patterns` · `dotnet-patterns` |\n| **Mobile** | `expo-router-navigation` · `flutter-patterns` |\n| **Security \u0026 Auth** | `owasp-top10` · `auth-nextauth` · `input-validation-sanitization` |\n| **DevOps \u0026 Cloud** | `github-actions-ci-cd` · `docker-containerization` · `vercel-deployment` · `aws-deployment` · `azure-deployment` · `gcp-deployment` |\n| **Performance** | `react-performance` · `caching-strategies` · `performance-testing` |\n| **Quality (QA)** | `vitest-unit-tests` · `playwright-e2e` · `react-testing-library` · `tdd-workflow` · `python-testing` |\n| **AI \u0026 Engineering** | `openai-sdk` · `anthropic-claude-sdk` · `google-gemini-sdk` · `vercel-ai-sdk` · `langchain-typescript` · `rag-implementation` |\n| **Marketing \u0026 SEO** | `seo-core-web-vitals` |\n| **UX/UI Design** | `ux-fundamentals` |\n\n\u003c/details\u003e\n\n---\n\n## ✅ Quality Gates (12)\n\nThe `studio validate` and `studio run` commands execute shell-based quality gates against your project. Gates are tiered by enforcement level:\n\n### Tier 1 — Hard Block (CI fails if these fail)\n| Gate | Purpose |\n|---|---|\n| `security-scan` | OWASP vulnerability patterns, secrets detection |\n| `ts-check` | TypeScript strict compilation — zero errors |\n| `env-validator` | `.env` schema validation — required keys present |\n\n### Tier 2 — Auto-Fix (applied automatically with `--fix`)\n| Gate | Purpose |\n|---|---|\n| `dependency-audit` | `npm audit` — patches non-breaking vulnerabilities |\n| `license-audit` | Detects GPL/AGPL licences in production dependencies |\n\n### Tier 3 — Advisory (warnings, does not block CI)\n| Gate | Purpose |\n|---|---|\n| `accessibility-audit` | WCAG 2.1 AA automated scan |\n| `bundle-analyzer` | Bundle size vs. budget thresholds |\n| `performance-budget` | Core Web Vitals — LCP, CLS, FID |\n| `seo-linter` | Missing meta tags, Open Graph, JSON-LD |\n| `i18n-linter` | Hardcoded strings not passed through i18n |\n| `type-coverage` | TypeScript coverage threshold (default 90%) |\n| `verify-all` | Runs all tiers in sequence |\n\nEach gate ships as `scripts/\u003cgate\u003e/\u003cstack\u003e.sh`, where `\u003cstack\u003e` is one of: `node` · `python` · `java` · `dotnet` · `flutter`.\n\n---\n\n## ⚡ Slash Commands (15)\n\nWorkflows are step-by-step AI procedures triggered by slash commands in your IDE:\n\n| Command | Purpose |\n|---|---|\n| `/blueprint` | Architecture doc for complex features — **required** before 3+ domain changes |\n| `/create` | Full feature scaffold (types → service → API → UI → tests) |\n| `/enhance` | Improve existing code quality, patterns, and performance |\n| `/debug` | Systematic root-cause analysis and fix |\n| `/audit-security` | Full OWASP security review of selected code |\n| `/refactor-solid` | Apply SOLID principles + Clean Architecture patterns |\n| `/generate-tests` | Unit + integration test suite generation |\n| `/generate-e2e` | Playwright end-to-end test generation |\n| `/deploy` | CI/CD pipeline and deployment configuration |\n| `/perf-audit` | React/Node performance profiling and optimisation |\n| `/a11y-audit` | Accessibility (WCAG 2.1 AA) compliance review |\n| `/document` | JSDoc, API docs, and README generation |\n| `/orchestrate` | Multi-agent coordination for epic-scale features |\n| `/preview` | Visual component and flow preview |\n| `/status` | Report installed agents, skills, and compliance posture |\n\n---\n\n## 💻 CLI Reference\n\n### `studio init`\nInitialize the Nexus agent system in your project.\n\n```bash\nstudio init\nstudio init --force          # Overwrite existing .agent/ files\nstudio init --path ./my-app  # Target a specific directory\nstudio init --quiet          # Suppress banner (for scripted use)\n```\n\n### `studio status`\nDisplay installed agents, skills, workflows, scripts, and enterprise compliance posture.\n\n```bash\nstudio status\n```\n\n### `studio list`\nBrowse the full registry: profiles, agents, skills, and slash commands.\n\n```bash\nstudio list\n```\n\n### `studio doctor`\nHealth check: Node.js version, `.agstudio.json` validity, `.agent/` structure, Git exclusions, company config, and registry integrity.\n\n```bash\nstudio doctor\n```\n\n### `studio validate`\nRun quality gate scripts against your project.\n\n```bash\nstudio validate\nstudio validate --fix         # Auto-fix linting/format issues\nstudio validate --skip-e2e    # Skip Playwright tests (faster CI)\nstudio validate --all         # Scan all monorepo packages (requires turbo)\n```\n\n### `studio run \u003cgate\u003e`\nRun a single quality gate by name.\n\n```bash\nstudio run security-scan\nstudio run dependency-audit --fix\nstudio run ts-check --stack node\nstudio run verify-all --dry-run\nstudio run --list             # Show all available gates\n```\n\n### `studio add \u003ctype\u003e \u003cid\u003e`\nAdd an individual agent, skill, workflow, or script to an already-initialised project.\n\n```bash\nstudio add agent  cloud-architect\nstudio add skill  owasp-top10\nstudio add workflow blueprint\n```\n\n### `studio remove \u003ctype\u003e \u003cid\u003e`\nRemove an installed component and update `.agstudio.json`.\n\n```bash\nstudio remove agent  seo-specialist\nstudio remove skill  framer-motion\n```\n\n### `studio update`\nCompare installed template hashes against the upstream registry and update outdated files.\n\n```bash\nstudio update\nstudio update --dry-run       # Preview what would change\nstudio update --force         # Overwrite including customised files\n```\n\n### `studio sync`\nDetect drift between installed files and the registry. Also enforces company skill policy.\n\n```bash\nstudio sync                   # Interactive: show drift, choose what to pull\nstudio sync --check           # CI mode — exit 1 if any drift detected\nstudio sync --force           # Pull all updates without prompting\n```\n\n### `studio company init \u003cname\u003e`\nScaffold a `.agstudio.company.json` to define org-wide required skills, forbidden patterns, and default IDE configs.\n\n```bash\nstudio company init \"ACME Corp\"\n```\n\n### `studio company validate`\nValidate a `.agstudio.company.json` against the schema.\n\n```bash\nstudio company validate\n```\n\n### `studio search \u003cquery\u003e`\nSearch the registry for agents, skills, workflows, and profiles by keyword.\n\n```bash\nstudio search \"security\"\nstudio search \"react\" --type skill\nstudio search \"java\"  --type agent\n```\n\n### `studio info \u003ctype\u003e \u003cid\u003e`\nShow the full definition, token budget, skill cross-references, and activation conditions for any registry item.\n\n```bash\nstudio info agent   enterprise-architect\nstudio info skill   owasp-top10\nstudio info profile nextjs-fullstack\n```\n\n### `studio diff`\nShow what has drifted between your installed files and the upstream registry (read-only — does not modify any files).\n\n```bash\nstudio diff                               # Show all drifted files\nstudio diff agents/security-engineer.md  # Diff a specific file\n```\n\n### `studio rollback`\nRoll back the most recent `update` or `sync` operation.\n\n```bash\nstudio rollback\nstudio rollback --dry-run   # Preview what would be restored\n```\n\n### `studio ci`\nCI-optimised single command: runs `validate` + `sync --check` in sequence. Exits non-zero on any failure — designed for pipeline use.\n\n```bash\nstudio ci\nstudio ci --strict          # Also fail on Tier 3 advisory warnings\n```\n\n### `studio create \u003ctype\u003e \u003cid\u003e`\nScaffold a new agent, skill, or workflow template locally — ready to customise and contribute upstream.\n\n```bash\nstudio create skill    my-graphql-patterns\nstudio create agent    my-domain-expert\nstudio create workflow my-deploy-flow\n```\n\n### `studio contribute \u003ctype\u003e \u003cid\u003e`\nValidate a locally-created template against the quality bar, auto-generate its registry entry, and print the git commands to open a PR.\n\n```bash\nstudio contribute skill  my-graphql-patterns\nstudio contribute agent  my-domain-expert\n```\n\n### `studio context`\nTrack AI context — decisions, architectural changes, and technical debt — linked to your project history.\n\n```bash\nstudio context init                           # Initialise context tracking\nstudio context sync                           # Sync to .agent/CONTEXT.md\nstudio context log \"Migrated auth to NextAuth v5\"\nstudio context status                         # Show entries + staleness\n```\n\n### `studio profile`\nSave and restore named sets of agents + skills for reuse across projects or team members.\n\n```bash\nstudio profile create  my-team-profile\nstudio profile show    my-team-profile\nstudio profile edit    my-team-profile\nstudio profile path                           # Print profile file location\n```\n\n### `studio mcp`\nManage MCP (Model Context Protocol) server integrations across all active IDE configs.\n\n```bash\nstudio mcp list                               # Show registered MCP servers\nstudio mcp add    @modelcontextprotocol/server-filesystem\nstudio mcp remove @modelcontextprotocol/server-filesystem\nstudio mcp apply                              # Write mcp config to all IDEs\n```\n\n### `studio completion`\nGenerate shell completion scripts for instant tab-completion of commands and IDs.\n\n```bash\nstudio completion bash  \u003e\u003e ~/.bashrc\nstudio completion zsh   \u003e\u003e ~/.zshrc\nstudio completion fish  \u003e ~/.config/fish/completions/studio.fish\n```\n\n---\n\n## 🏢 Enterprise Configuration\n\nFor organisations deploying Nexus Studio across multiple teams, create a `.agstudio.company.json`:\n\n```json\n{\n  \"companyName\": \"ACME Corp\",\n  \"version\": \"1.0.0\",\n  \"requiredSkills\": [\"owasp-top10\", \"clean-architecture\", \"solid-principles\"],\n  \"forbiddenSkills\": [],\n  \"defaultIdes\": [\"copilot\"],\n  \"codingStandardsUrl\": \"https://wiki.acme.com/engineering-standards\",\n  \"registryUrl\": \"https://registry.acme.com/nexus\"\n}\n```\n\n**How it works:**\n- Place in project root (per-repo) or `~/.agstudio.company.json` (machine-wide via IT provisioning)\n- `studio init` automatically detects it and injects required skills, removes forbidden ones\n- `studio sync` enforces the policy on every drift check\n- `studio status` shows a compliance percentage bar\n\n---\n\n## 🖥️ IDE Integration\n\nWhen you run `studio init`, Nexus generates IDE-specific config files that bridge your IDE's AI assistant to the `.agent/` system:\n\n| IDE | Files Generated |\n|---|---|\n| **GitHub Copilot** | `.github/copilot-instructions.md` · `.github/prompts/*.prompt.md` · `.github/instructions/agents.instructions.md` · `.vscode/settings.json` · `.vscode/mcp.json` |\n| **Cursor** | `.cursor/rules/nexus-system.mdc` · `.cursor/rules/{agent}.mdc` · `.cursor/mcp.json` · `.cursorignore` |\n| **Windsurf** | `.windsurfrules` · `.codeiumignore` |\n| **Claude Code** | `CLAUDE.md` · `.claude/settings.json` · `.claude/commands/*.md` · `.claude/agents/*.md` · `.mcp.json` |\n\n\u003e `.agent/` is always installed regardless of IDE — it is the source of truth that all IDE files are generated from.\n\n---\n\n## 🗂️ Project Configuration (`.agstudio.json`)\n\nAfter `studio init`, a `.agstudio.json` is created in your project root:\n\n```json\n{\n  \"version\": \"1.0.0\",\n  \"profile\": \"nextjs-fullstack\",\n  \"project\": \"my-app\",\n  \"installed\": {\n    \"agents\": [\"enterprise-architect\", \"nextjs-expert\", \"security-engineer\"],\n    \"skills\": [\"nextjs-app-router\", \"owasp-top10\", \"clean-architecture\"],\n    \"workflows\": [\"blueprint\", \"create\", \"debug\"],\n    \"scripts\": [\"security-scan\", \"ts-check\", \"verify-all\"]\n  },\n  \"customized\": [],\n  \"telemetry\": false,\n  \"installedHashes\": {\n    \"agents/enterprise-architect.md\": \"a1b2c3d4e5f6\"\n  }\n}\n```\n\n`installedHashes` stores SHA-256 fingerprints of each installed template at install time. `studio sync` and `studio update` compare these against the current upstream templates to detect drift without false positives from Handlebars compilation.\n\n---\n\n## 🔒 Security \u0026 Git\n\n`.agent/` is **local-only** by design. `studio init` automatically adds it to `.git/info/exclude` (not `.gitignore`) so:\n- ✅ The AI can index and read `.agent/` files\n- ✅ The files never appear in `git status`\n- ✅ The files are never committed or pushed\n\n\u003e **Do not add `.agent/` to `.gitignore`** — this breaks IDE indexing. Use `.git/info/exclude` (which `studio init` handles automatically).\n\n---\n\n## 🧪 Testing\n\n```bash\nnpm test          # Run all 43 tests via Vitest + Turbo\nnpm run typecheck # TypeScript strict check — 0 errors\nnpm run build     # Production build via tsup\n```\n\n**Test coverage:**\n\n| Suite | Tests |\n|---|---|\n| `platform.test.ts` | 4 — cross-platform bash resolver |\n| `config-manager.test.ts` | 5 — config CRUD + hash persistence |\n| `project-detector.test.ts` | 8 — framework + IDE heuristics |\n| `enterprise-config.test.ts` | 9 — company policy load/validate/apply |\n| `add-remove.test.ts` | 7 — input sanitisation (path traversal, injection) |\n| `status.test.ts` | 2 — command happy path + error state |\n| `template-engine.test.ts` | 8 — Handlebars safe compile (JSX, GH Actions, block helpers) |\n| **Total** | **43 passed** |\n\n---\n\n## 📁 Monorepo Structure\n\n```\nnexus-studio-monorepo/\n├── apps/\n│   └── studio/              # @nexus/studio — the CLI package\n│       ├── src/\n│       │   ├── cli.ts       # Commander.js entry point (21 command modules)\n│       │   ├── commands/    # init · status · list · doctor · validate · run\n│       │   │                # add · remove · update · sync · company · search\n│       │   │                # info · diff · rollback · ci · create · contribute\n│       │   │                # context · profile · mcp · completion\n│       │   ├── core/        # template-engine · project-detector · config-manager\n│       │   │                # ide-config-generator · enterprise-config · git-integration\n│       │   ├── ui/          # banner · logger · spinner\n│       │   └── types/       # config · registry · validation\n│       └── templates/       # (symlinked from packages/templates at build time)\n├── packages/\n│   └── templates/           # @nexus/templates — registry + agent/skill/workflow files\n│       ├── registry.json    # Single source of truth for all agents, skills, profiles\n│       ├── .agent/          # Template files (Handlebars — compiled at install time)\n│       ├── .cursor/         # Cursor IDE template rules\n│       ├── .claude/         # Claude Code template (CLAUDE.md)\n│       ├── .github/         # Copilot template (copilot-instructions.md)\n│       └── windsurf/        # Windsurf template (.windsurfrules)\n└── scripts/                 # Maintenance scripts (standardize-skills · compute-token-budgets)\n```\n\n---\n\n## 📋 Requirements\n\n| Requirement | Minimum |\n|---|---|\n| Node.js | `\u003e= 18.0.0` |\n| npm | `\u003e= 10.0.0` |\n| OS | Windows · macOS · Linux |\n| Git | Any recent version (for `.git/info/exclude` setup) |\n\n\u003e **Windows users**: Quality gate scripts (`.sh`) require Git Bash. Nexus Studio auto-detects Git Bash at the standard install paths. WSL is also supported.\n\n---\n\n## 🤝 Contributing\n\nContributions are welcome — and the fastest path is via the CLI itself:\n\n```bash\n# 1. Scaffold a new template\nstudio create skill  my-graphql-patterns\n\n# 2. Edit the generated file\n#    packages/templates/.agent/skills/my-graphql-patterns/SKILL.md\n\n# 3. Validate + generate registry entry + print PR git commands\nstudio contribute skill  my-graphql-patterns\n```\n\n**Quality bar per template type:**\n\n| Type | Minimum requirements |\n|---|---|\n| **Skill** | ≥ 400 chars · one code example · anti-patterns list · checklist · accurate `tokenBudget` |\n| **Agent** | Activation condition · when NOT to activate · output format spec · skill cross-references |\n| **Workflow** | Trigger command · step-by-step · success criteria · rollback instructions |\n\n**Review turnaround:**\n\n| PR type | Reviewer | ETA |\n|---|---|---|\n| New skill | 1 senior dev | ~2 days |\n| New agent | Deepak Biswal | ~1 week |\n| New workflow | 1 senior dev + QA | ~3 days |\n| Bug fix | Automated CI | Same day |\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md) for the complete YAML frontmatter schema, branch naming convention (`feat/contribute-\u003ctype\u003e-\u003cid\u003e`), and test requirements.\n\n---\n\n## 📄 License\n\nMIT © Deepak Biswal\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdeepakb%2Fantigravity-studio","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdeepakb%2Fantigravity-studio","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdeepakb%2Fantigravity-studio/lists"}