\n Human-in-the-loop approval system for AI agents. \n Agents request. Policies decide. Humans approve. \n Keep humans in control of what AI agents can do.\n
\n\n\n
\n \n \n \n \n
\n\n---\n\n
\n \n
\n\nYour AI agent wants to send an email, delete a file, or deploy to production.\n**Should it?** AgentGate lets you define policies that auto-approve safe actions,\nauto-deny dangerous ones, and route everything else to a human β via dashboard, Slack, Discord, or email.\n\n### β¨ Highlights\n\n- π‘οΈ **Policy engine** β auto-approve, auto-deny, or route to humans based on rules\n- π₯ **Multi-channel approvals** β Slack, Discord, email, or web dashboard\n- π **TypeScript SDK + MCP** β works with any agent framework or Claude Desktop\n- πͺ **Webhooks with retry** β real-time notifications with exponential backoff\n- π **Full audit trail** β every request, decision, and action logged, each policy decision tagged with the **OWASP LLM Top-10** risk it mitigates (compliance evidence)\n- π³ **Docker-ready** β one `docker-compose up` for the full stack\n- π **Production-hardened** β SSRF protection, ReDoS defense, structured logging, graceful shutdown\n- π **One-click decision links** β approve or deny directly from notification emails and webhooks\n- βΏ **Accessible UI** β keyboard-navigable approval modals, focus trapping, ARIA labels\n- π **Skeleton loading** β smooth loading states across every dashboard page\n- β‘ **Fast & lightweight** β Hono server, SQLite or PostgreSQL\n\n## Quickstart\n\n```bash\n# Self-host the full stack (server + dashboard + Postgres):\ndocker compose up\n\n# β¦or from source:\npnpm install && pnpm --filter @agentkitai/agentgate-server db:migrate && pnpm --filter @agentkitai/agentgate-server bootstrap && pnpm dev\n```\n\nDrop AgentGate into an MCP client (Claude Desktop / Cursor / VS Code) β point it at the gateway:\n\n```jsonc\n{ \"mcpServers\": { \"agentgate\": { \"command\": \"npx\", \"args\": [\"@agentkitai/agentgate-mcp\"] } } }\n```\n\n#### Dashboard\nSee all pending requests at a glance, color-coded by urgency so you know what needs attention first.\n\n\n\n#### Approval Requests\nReview, approve, or deny requests β filter by status to focus on what matters.\n\n\n\n#### Audit Log\nSearch through every decision with filters for event type, action, actor, and date range.\n\n\n\n#### Request Detail\nDrill into any request to see parameters, context, timeline, and audit trail β with one-click Approve/Deny buttons.\n\n\n\n#### API Keys\nManage API keys with fine-grained scopes, rate limits, and usage tracking. Create, edit, or revoke keys from the dashboard.\n\n\n\n#### Webhooks\nConfigure webhook endpoints for real-time notifications. Add URLs, pick events, and let AgentGate handle retries automatically.\n\n\n\n#### Login\nSign in with your API key β create one via the CLI or ask your admin.\n\n\n\n---\n\n## Table of Contents\n\n- [Quick Start](#quick-start)\n- [Architecture](#architecture)\n- [Packages](#packages)\n- [SDK Usage](#sdk-usage)\n- [CLI](#cli)\n- [MCP Integration](#mcp-integration)\n- [Authentication](#authentication)\n- [API Endpoints](#api-endpoints)\n- [Rate Limiting](#rate-limiting)\n- [Webhooks](#webhooks)\n- [Configuration](#configuration)\n- [Docker Deployment](#docker-deployment)\n- [Development](#development)\n- [Troubleshooting](docs/troubleshooting.md)\n- [Contributing](#contributing)\n- [License](#license)\n\n---\n\n## Quick Start\n\n### 1. Install dependencies\n\n```bash\npnpm install\n```\n\n### 2. Run database migrations\n\n```bash\npnpm --filter @agentkitai/agentgate-server db:migrate\n```\n\n### 3. Bootstrap (create admin API key)\n\n```bash\npnpm --filter @agentkitai/agentgate-server bootstrap\n```\n\n**Save the API key** - it's shown once only! Set it in your environment:\n\n```bash\nexport AGENTGATE_API_KEY=\"agk_...\"\n```\n\n### 4. Start the development environment\n\n```bash\n# Start server (port 3000) and dashboard (port 5173)\npnpm dev\n```\n\n### 5. Run the demo\n\nIn a new terminal (with API key set):\n\n```bash\nexport AGENTGATE_API_KEY=\"agk_...\"\npnpm demo\n```\n\n### 6. Open the dashboard\n\nVisit **http://localhost:5173** to view and manage approval requests.\n\n## Architecture\n\n```\nβββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ\nβ AI Agents β\nβ (use @agentkitai/agentgate-sdk or MCP to request approvals) β\nβββββββββββββββββββββββββββββ¬ββββββββββββββββββββββββββββββββββββββ\n β HTTP API (authenticated)\n βΌ\nβββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ\nβ AgentGate Server β\nβ ββββββββββββββββ ββββββββββββββββ ββββββββββββββββ β\nβ β Policy Engineβ β Request Storeβ β Audit Logger β β\nβ ββββββββββββββββ€ ββββββββββββββββ€ ββββββββββββββββ€ β\nβ β API Keys β β Webhooks β β MCP Server β β\nβ ββββββββββββββββ ββββββββββββββββ ββββββββββββββββ β\nβββββββββββββββββββββββββββββ¬ββββββββββββββββββββββββββββββββββββββ\n β\n βββββββββββββββΌββββββββββββββ\n βΌ βΌ βΌ\nββββββββββββββββββ ββββββββββββββββββ ββββββββββββββββββ\nβ Web Dashboard β β Slack Bot β β Discord Bot β\nβ(React+Tailwind)β β(approve in DM) β β(approve in ch) β\nββββββββββββββββββ ββββββββββββββββββ ββββββββββββββββββ\n β β β\n βββββββββββββββΌββββββββββββββ\n βΌ\n ββββββββββββ\n β Humans β\n ββββββββββββ\n```\n\n## Packages\n\n| Package | Description | Docs |\n|---------|-------------|------|\n| [`@agentkitai/agentgate-core`](./packages/core) | Types, schemas, policy engine | - |\n| [`@agentkitai/agentgate-server`](./packages/server) | Hono API server | - |\n| [`@agentkitai/agentgate-sdk`](./packages/sdk) | TypeScript SDK for agents | [README](./packages/sdk/README.md) |\n| [`@agentkitai/agentgate-cli`](./packages/cli) | Command-line interface | - |\n| [`@agentkitai/agentgate-mcp`](./packages/mcp) | MCP server for Claude Desktop | - |\n| [`@agentkitai/agentgate-slack`](./packages/slack) | Slack bot integration | [README](./packages/slack/README.md) |\n| [`@agentkitai/agentgate-discord`](./packages/discord) | Discord bot integration | [README](./packages/discord/README.md) |\n| [`@agentkitai/agentgate-dashboard`](./packages/dashboard) | React web dashboard | - |\n\n## SDK Usage\n\n```typescript\nimport { AgentGateClient } from '@agentkitai/agentgate-sdk';\n\n// Create client with API key\nconst client = new AgentGateClient({\n baseUrl: 'http://localhost:3000',\n apiKey: process.env.AGENTGATE_API_KEY,\n});\n\n// Request approval\nconst request = await client.request({\n action: 'send_email',\n params: {\n to: 'customer@example.com',\n subject: 'Order shipped!',\n },\n urgency: 'normal',\n});\n\n// Wait for human decision\nconst decided = await client.waitForDecision(request.id, {\n timeout: 60000, // 1 minute\n});\n\nif (decided.status === 'approved') {\n // Execute the action\n await sendEmail(decided.params);\n} else {\n console.log('Action denied:', decided.decisionReason);\n}\n```\n\n## CLI\n\nAgentGate includes a command-line interface for managing approval requests.\n\n### Installation\n\n```bash\n# From the monorepo\npnpm --filter @agentkitai/agentgate-cli build\n\n# Or install globally (when published)\nnpm install -g @agentkitai/agentgate-cli\n```\n\n### Configuration\n\nConfigure the CLI with your server URL and API key:\n\n```bash\n# Set server URL\nagentgate config set serverUrl http://localhost:3000\n\n# Set API key\nagentgate config set apiKey agk_your_api_key\n\n# View current config\nagentgate config show\n```\n\nConfiguration is stored in `~/.agentgate/config.json`. You can also use environment variables:\n\n```bash\nexport AGENTGATE_URL=http://localhost:3000\nexport AGENTGATE_API_KEY=agk_...\n```\n\n### Commands\n\n| Command | Description |\n|---------|-------------|\n| `agentgate config show` | Show current configuration |\n| `agentgate config set ` | Set a configuration value |\n| `agentgate request ` | Create a new approval request |\n| `agentgate status ` | Get status of a request |\n| `agentgate list` | List approval requests |\n| `agentgate approve ` | Approve a pending request |\n| `agentgate deny ` | Deny a pending request |\n\n### Examples\n\n```bash\n# Create a request\nagentgate request send_email \\\n --params '{\"to\": \"user@example.com\", \"subject\": \"Hello\"}' \\\n --urgency high\n\n# List pending requests\nagentgate list --status pending\n\n# Approve a request\nagentgate approve req_abc123 --reason \"Looks good\"\n\n# Deny a request\nagentgate deny req_abc123 --reason \"Not authorized\"\n\n# Output as JSON\nagentgate list --json\n```\n\n## MCP Integration\n\nAgentGate includes a Model Context Protocol (MCP) server for integration with Claude Desktop and other MCP-compatible clients.\n\n### Claude Desktop Configuration\n\nAdd to your `claude_desktop_config.json`:\n\n```json\n{\n \"mcpServers\": {\n \"agentgate\": {\n \"command\": \"npx\",\n \"args\": [\"@agentkitai/agentgate-mcp\"],\n \"env\": {\n \"AGENTGATE_URL\": \"http://localhost:3000\",\n \"AGENTGATE_API_KEY\": \"agk_...\"\n }\n }\n }\n}\n```\n\n### Available MCP Tools\n\n| Tool | Description |\n|------|-------------|\n| `agentgate_request` | Submit a new approval request |\n| `agentgate_get` | Get the status of an approval request by ID |\n| `agentgate_list` | List approval requests with optional filters |\n| `agentgate_decide` | Approve or deny a pending request |\n| `agentgate_list_policies` | List all policies ordered by priority |\n| `agentgate_create_policy` | Create a new policy with rules |\n| `agentgate_update_policy` | Replace an existing policy |\n| `agentgate_delete_policy` | Delete a policy by ID |\n| `agentgate_list_audit_logs` | List audit log entries with filters and pagination |\n| `agentgate_get_audit_actors` | Get unique actor values from audit logs |\n\n## Authentication\n\nAgentGate uses API keys for authentication. All API requests (except `/health`) require a valid API key.\n\n### API Key Scopes\n\n| Scope | Description |\n|-------|-------------|\n| `admin` | Full access to all operations |\n| `request:create` | Create new approval requests |\n| `request:read` | Read approval requests |\n| `request:decide` | Approve or deny requests |\n| `webhook:manage` | Create/update/delete webhooks |\n\n### Using API Keys\n\n**HTTP Header:**\n```bash\ncurl -H \"Authorization: Bearer agk_...\" http://localhost:3000/api/requests\n```\n\n**SDK:**\n```typescript\nconst client = new AgentGateClient({\n baseUrl: 'http://localhost:3000',\n apiKey: process.env.AGENTGATE_API_KEY,\n});\n```\n\n### Creating Additional API Keys\n\n```typescript\n// Via API (requires admin scope)\nPOST /api/api-keys\n{\n \"name\": \"My Agent\",\n \"scopes\": [\"request:create\", \"request:read\"]\n}\n```\n\n## API Endpoints\n\n| Method | Endpoint | Description | Required Scope |\n|--------|----------|-------------|----------------|\n| `POST` | `/api/requests` | Create approval request | `request:create` |\n| `GET` | `/api/requests` | List requests (with filters) | `request:read` |\n| `GET` | `/api/requests/:id` | Get request by ID | `request:read` |\n| `POST` | `/api/requests/:id/decide` | Submit approval/denial | `request:decide` |\n| `GET` | `/api/requests/:id/audit` | Get audit trail | `request:read` |\n| `GET` | `/api/policies` | List policies | `admin` |\n| `POST` | `/api/policies` | Create policy | `admin` |\n| `PUT` | `/api/policies/:id` | Update policy | `admin` |\n| `DELETE` | `/api/policies/:id` | Delete policy | `admin` |\n| `POST` | `/api/api-keys` | Create API key | `admin` |\n| `GET` | `/api/api-keys` | List API keys | `admin` |\n| `PATCH` | `/api/api-keys/:id` | Update API key | `admin` |\n| `DELETE` | `/api/api-keys/:id` | Revoke API key | `admin` |\n| `GET` | `/api/webhooks` | List webhooks | `webhook:manage` |\n| `POST` | `/api/webhooks` | Create webhook | `webhook:manage` |\n| `DELETE` | `/api/webhooks/:id` | Delete webhook | `webhook:manage` |\n| `GET` | `/health` | Health check | (none) |\n\n## Rate Limiting\n\nAgentGate supports per-API-key rate limiting to prevent abuse and ensure fair usage.\n\n### How It Works\n\n- Rate limits use a **sliding window algorithm** (requests per minute)\n- Limits are configured per API key\n- When exceeded, requests return `429 Too Many Requests`\n- Rate limit headers are included in all authenticated responses\n\n### Rate Limit Headers\n\n| Header | Description |\n|--------|-------------|\n| `X-RateLimit-Limit` | Maximum requests per minute |\n| `X-RateLimit-Remaining` | Remaining requests in current window |\n| `X-RateLimit-Reset` | Seconds until window resets |\n\n### Configuring Rate Limits\n\nSet rate limits when creating or updating API keys:\n\n```typescript\n// Via API (requires admin scope)\nPOST /api/api-keys\n{\n \"name\": \"My Agent\",\n \"scopes\": [\"request:create\", \"request:read\"],\n \"rateLimit\": 60 // 60 requests per minute\n}\n\n// null = unlimited\n{\n \"name\": \"Internal Service\",\n \"scopes\": [\"admin\"],\n \"rateLimit\": null\n}\n```\n\n### Dashboard\n\nRate limits can also be managed from the web dashboard under **Settings β API Keys**.\n\n## Webhooks\n\nAgentGate can notify external systems when request events occur.\n\n### Setting Up Webhooks\n\n```typescript\n// Create a webhook via API\nPOST /api/webhooks\n{\n \"url\": \"https://your-server.com/webhook\",\n \"events\": [\"request.created\", \"request.decided\"],\n \"secret\": \"optional-signing-secret\"\n}\n```\n\n### Webhook Events\n\n| Event | Description |\n|-------|-------------|\n| `request.created` | A new approval request was created |\n| `request.decided` | A request was approved or denied |\n| `request.expired` | A request expired without decision |\n\n### Webhook Payload\n\n```json\n{\n \"event\": \"request.decided\",\n \"timestamp\": \"2024-01-15T10:30:00Z\",\n \"data\": {\n \"id\": \"abc123\",\n \"action\": \"send_email\",\n \"status\": \"approved\",\n \"decidedBy\": \"admin@example.com\"\n }\n}\n```\n\n### Webhook Signatures\n\nIf you provide a `secret`, requests are signed with HMAC-SHA256:\n\n```\nX-AgentGate-Signature: sha256=...\n```\n\nVerify by computing `HMAC-SHA256(secret, body)` and comparing.\n\n### Webhook Retry\n\nFailed webhook deliveries are retried automatically with exponential backoff. The server scans for pending deliveries and retries them with increasing delays (`2^attempts * 1000ms`) until successful or the maximum retry count is reached.\n\n## Configuration\n\n### Environment Variables\n\n| Variable | Default | Description |\n|----------|---------|-------------|\n| `PORT` | `3000` | Server port |\n| `DATABASE_URL` | `./data/agentgate.db` | SQLite database path |\n| `AGENTGATE_API_KEY` | - | API key for SDK/CLI |\n| `SLACK_BOT_TOKEN` | - | Slack bot token (for Slack integration) |\n| `SLACK_SIGNING_SECRET` | - | Slack signing secret |\n| `DISCORD_BOT_TOKEN` | - | Discord bot token (for Discord integration) |\n| `DISCORD_DEFAULT_CHANNEL` | - | Default Discord channel for notifications |\n\n### File-Based Secrets (`_FILE` suffix)\n\nFor Docker secrets or Kubernetes secret mounts, AgentGate supports a `_FILE` suffix convention. Instead of setting a secret directly in an environment variable, point to a file containing the value:\n\n| Variable | Reads secret from file |\n|----------|----------------------|\n| `ADMIN_API_KEY_FILE` | Sets `ADMIN_API_KEY` |\n| `JWT_SECRET_FILE` | Sets `JWT_SECRET` |\n| `DATABASE_URL_FILE` | Sets `DATABASE_URL` |\n| `REDIS_URL_FILE` | Sets `REDIS_URL` |\n| `SLACK_BOT_TOKEN_FILE` | Sets `SLACK_BOT_TOKEN` |\n| `SLACK_SIGNING_SECRET_FILE` | Sets `SLACK_SIGNING_SECRET` |\n| `DISCORD_BOT_TOKEN_FILE` | Sets `DISCORD_BOT_TOKEN` |\n| `SMTP_PASS_FILE` | Sets `SMTP_PASS` |\n\n**Behavior:**\n- File contents are trimmed of leading/trailing whitespace\n- If both the env var and the `_FILE` variant are set, the explicit env var takes precedence\n- Missing or unreadable files produce a warning but do not crash the server\n\nExample with Docker Compose:\n\n```yaml\nservices:\n agentgate:\n environment:\n ADMIN_API_KEY_FILE: /run/secrets/admin_api_key\n JWT_SECRET_FILE: /run/secrets/jwt_secret\n secrets:\n - admin_api_key\n - jwt_secret\n\nsecrets:\n admin_api_key:\n file: ./secrets/admin_api_key.txt\n jwt_secret:\n file: ./secrets/jwt_secret.txt\n```\n\n### Policy Configuration\n\nPolicies are stored in the database and can be managed via API:\n\n```typescript\n// Example: Auto-approve low-risk emails\n{\n name: \"auto-approve-emails\",\n priority: 10,\n enabled: true,\n rules: [\n {\n match: { action: \"send_email\" },\n decision: \"auto_approve\"\n }\n ]\n}\n```\n\n## Docker Deployment\n\nAgentGate provides Docker images for easy self-hosted deployments.\n\n### Quick Start\n\n1. **Copy the example environment file:**\n\n```bash\ncp .env.example .env\n```\n\n2. **Generate secure credentials:**\n\n```bash\n# Generate admin API key (required)\necho \"ADMIN_API_KEY=$(openssl rand -hex 32)\" >> .env\n\n# Generate JWT secret (recommended for production)\necho \"JWT_SECRET=$(openssl rand -hex 32)\" >> .env\n```\n\n3. **Start all services:**\n\n```bash\ndocker-compose up -d\n```\n\n4. **Access the services:**\n\n- **Dashboard:** http://localhost:3003\n- **API Server:** http://localhost:3002\n- **Health Check:** http://localhost:3002/health\n\n> Host ports are configurable via `SERVER_PORT` (default `3002`) and `DASHBOARD_PORT` (default `3003`); both map to the container's internal port 3000/80.\n\n### Services\n\n| Service | Description | Host Port |\n|---------|-------------|------|\n| `server` | AgentGate API server | 3002 |\n| `dashboard` | Web dashboard (nginx) | 3003 |\n| `postgres` | PostgreSQL database | internal only* |\n| `redis` | Redis (rate limiting, queues) | internal only* |\n\n> \\* PostgreSQL and Redis are on an internal Docker network (`agentgate-internal`) and are **not exposed to the host** by default. During development, `docker-compose.override.yml` is auto-loaded and exposes them on ports 5432/6379. For production, run `docker-compose -f docker-compose.yml up -d` to skip the override.\n\n### With Slack or Discord Bots\n\nTo include the bot services, use the `bots` profile:\n\n```bash\n# Set required bot credentials in .env first\ndocker-compose --profile bots up -d\n```\n\n### Configuration\n\nAll configuration is done via environment variables. See `.env.example` for all options.\n\n**Required variables:**\n- `ADMIN_API_KEY` β Admin API key (min 16 characters)\n\n**Recommended for production:**\n- `JWT_SECRET` β JWT signing secret (min 32 characters)\n- `CORS_ALLOWED_ORIGINS` β Restrict to your domain(s)\n- `POSTGRES_PASSWORD` β Use a strong password\n\n### Building Images\n\nBuild all images locally:\n\n```bash\ndocker-compose build\n```\n\nBuild a specific service:\n\n```bash\ndocker-compose build server\ndocker-compose build dashboard\n```\n\n### Database Migrations\n\nMigrations run automatically when the server starts. For manual control:\n\n```bash\n# Run migrations inside the container\ndocker-compose exec server node -e \"\n import('./dist/db/migrate.js').then(m => m.runMigrations())\n\"\n```\n\n### Viewing Logs\n\n```bash\n# All services\ndocker-compose logs -f\n\n# Specific service\ndocker-compose logs -f server\n\n# Last 100 lines\ndocker-compose logs --tail=100 server\n```\n\n### Stopping Services\n\n```bash\n# Stop all\ndocker-compose down\n\n# Stop and remove volumes (WARNING: deletes data)\ndocker-compose down -v\n```\n\n### Production Considerations\n\n1. **Use a reverse proxy** (nginx, Caddy, Traefik) for TLS termination\n2. **Set strong passwords** for PostgreSQL\n3. **Restrict CORS origins** to your domain\n4. **Use Docker secrets** for sensitive values in production\n5. **Set up backups** for PostgreSQL data volume\n6. **Monitor health endpoints** for uptime checks\n\n## Development\n\n```bash\n# Install dependencies\npnpm install\n\n# Run migrations\npnpm --filter @agentkitai/agentgate-server db:migrate\n\n# Bootstrap (create admin key)\npnpm --filter @agentkitai/agentgate-server bootstrap\n\n# Start development (server + dashboard)\npnpm dev\n```\n\n### Testing\n\nAgentGate uses [Vitest](https://vitest.dev/) for testing across all packages.\n\n```bash\n# Run all tests\npnpm test\n\n# Run tests with coverage report\npnpm test:coverage\n\n# Run tests in watch mode (single package)\npnpm --filter @agentkitai/agentgate-server test:watch\n\n# Run a specific test file\npnpm --filter @agentkitai/agentgate-server test -- src/__tests__/integration.test.ts\n```\n\nCoverage reports are generated per-package and include line, branch, and function coverage.\n\n### Code Quality\n\n```bash\n# Build all packages\npnpm build\n\n# Type checking\npnpm typecheck\n\n# Lint (ESLint)\npnpm lint\n\n# Fix lint issues\npnpm lint:fix\n\n# Format code (Prettier)\npnpm format\n\n# Check formatting\npnpm format:check\n```\n\n## Project Structure\n\n```\nagentgate/\nβββ packages/\nβ βββ core/ # Shared types, schemas, policy engine\nβ βββ server/ # Hono API server\nβ βββ sdk/ # TypeScript SDK\nβ βββ cli/ # Command-line interface\nβ βββ mcp/ # MCP server for Claude Desktop\nβ βββ slack/ # Slack bot\nβ βββ discord/ # Discord bot\nβ βββ dashboard/ # React dashboard\nβββ apps/\nβ βββ demo/ # Demo application\nβββ docker-compose.yml # Docker deployment\nβββ package.json # Monorepo root\n```\n\n## Contributing\n\nContributions are welcome! To get started:\n\n1. Fork the repository\n2. Clone and install dependencies (`pnpm install`)\n3. Follow the [Development](#development) section above to set up your local environment\n4. Create a feature branch and make your changes\n5. Run `pnpm build && pnpm test` to verify everything works\n6. Open a pull request\n\nPlease make sure all tests pass and code is formatted (`pnpm format:check && pnpm lint`) before submitting.\n\n\n## π§° AgentKit Ecosystem\n\n| Project | Description | |\n|---------|-------------|-|\n| [AgentLens](https://github.com/agentkitai/agentlens) | Observability & audit trail for AI agents | |\n| [Lore](https://github.com/agentkitai/lore) | Cross-agent memory and lesson sharing | |\n| **AgentGate** | Human-in-the-loop approval gateway | β¬ οΈ you are here |\n| [FormBridge](https://github.com/agentkitai/formbridge) | Agent-human mixed-mode forms | |\n| [AgentEval](https://github.com/agentkitai/agenteval) | Testing & evaluation framework | |\n| [agentkit-cli](https://github.com/agentkitai/agentkit-cli) | Unified CLI orchestrator | |\n\n## License\n\nMIT\n",
"bytes": 23267,
"sha": "0e3b182ddb94742e2641ebf1f65b76c939ca04e9241ba5e671c0aeb0653fd6f8",
"repo_slug": "agentkitai/agentgate",
"fonte": "repo",
"truncated": false,
"api": "https://agentalog.com/api/listings/mcp_io_github_agentkitai_agentgate_124f5646/readme"
}
![\"npm](\"https://img.shields.io/npm/v/@agentkitai/agentgate-sdk?label=sdk&color=blue\")
\n ![\"MIT](\"https://img.shields.io/badge/license-MIT-green\"){kind=icon}
\n ![\"CI\"](\"https://img.shields.io/github/actions/workflow/status/agentkitai/agentgate/ci.yml?branch=main\")
\n ![\"TypeScript\"](\"https://img.shields.io/badge/TypeScript-5.x-blue?logo=typescript&logoColor=white\"){kind=icon}
\n![\"AgentGate](\"docs/public/demo.gif\")
\n