{
  "markdown": "# Declaw MCP Server\n\nMCP server for [Declaw](https://declaw.ai) — secure sandbox execution for AI agents with network policies, PII scanning, prompt injection defense, and audit logging.\n\nWorks with Claude Desktop, Claude Code, Cursor, Windsurf, and any MCP-compatible AI tool.\n\n## Quick Start\n\n### Claude Desktop / Cursor / Windsurf\n\nAdd to your MCP config:\n\n```json\n{\n  \"mcpServers\": {\n    \"declaw\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"@declaw/mcp-server\"],\n      \"env\": {\n        \"DECLAW_API_KEY\": \"your-api-key\"\n      }\n    }\n  }\n}\n```\n\n### Claude Code\n\n```bash\nclaude mcp add declaw -- npx -y @declaw/mcp-server\n```\n\nSet `DECLAW_API_KEY` in your environment.\n\n## Tools\n\n| Tool | Description |\n|------|-------------|\n| `create_sandbox` | Create a secure sandbox with configurable security policies |\n| `run_command` | Execute a shell command inside a sandbox |\n| `read_file` | Read a file from a sandbox |\n| `write_file` | Write a file to a sandbox |\n| `list_files` | List directory contents in a sandbox |\n| `kill_sandbox` | Destroy a sandbox |\n| `list_sandboxes` | List all active sandboxes |\n\n## Security Presets\n\nWhen creating a sandbox, choose a security preset:\n\n- **`none`** — No guardrails. Full internet access.\n- **`standard`** (default) — PII scanning + audit logging. Full internet access.\n- **`strict`** — PII scanning + prompt injection defense + audit logging + network deny-all.\n\nYou can also pass `allowed_domains` to restrict outbound traffic to specific domains:\n\n```\ncreate_sandbox with template=\"python\", security_preset=\"strict\", allowed_domains=[\"pypi.org\", \"github.com\"]\n```\n\n## Why Declaw?\n\n| | Declaw | Other Sandbox Providers |\n|---|---|---|\n| Sandbox execution | Yes | Yes |\n| Non-bypassable network controls | Yes | ?? |\n| PII scanning | Yes | No |\n| Injection defense | Yes | No |\n| Full audit trail | Yes | Basic |\n| Snapshots | Yes | Varies |\n| Multiple templates | 8 built-in | Varies |\n| Interactive stdio | Yes | Varies |\n\n## Environment Variables\n\n| Variable | Required | Description |\n|----------|----------|-------------|\n| `DECLAW_API_KEY` | Yes | Your Declaw API key |\n| `DECLAW_DOMAIN` | No | Custom API domain (for on-prem deployments) |\n\n## On-Prem\n\nFor self-hosted Declaw deployments, set the domain:\n\n```json\n{\n  \"mcpServers\": {\n    \"declaw\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"@declaw/mcp-server\"],\n      \"env\": {\n        \"DECLAW_API_KEY\": \"your-api-key\",\n        \"DECLAW_DOMAIN\": \"declaw.internal.company.com\"\n      }\n    }\n  }\n}\n```\n\n## License\n\nApache-2.0\n",
  "bytes": 2534,
  "sha": "53ee7940ce4a37fe96f157f82a448131cd93242b2541b1a166e6db83254bc95b",
  "repo_slug": "declaw-ai/mcp-server",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_declaw_ai_mcp_server_9dc7601c/readme"
}