Safety and execution limits — AgentTwin Documentation MCP Resource
Safety and execution limits
is one of 35 resources on the
AgentTwin Documentation
MCP server. Connect the server and your client can attach it as context.
by client
How to attach Safety and execution limits from your client
- Claude Code AgentTwin Documentation Resource/Safety and execution limits run in your project directory
- Claude Desktop AgentTwin Documentation Resource/Safety and execution limits ~/Library/Application Support/Claude/claude_desktop_config.json
- Cursor AgentTwin Documentation Resource/Safety and execution limits ~/.cursor/mcp.json
- VS Code AgentTwin Documentation Resource/Safety and execution limits .vscode/mcp.json
- Zed AgentTwin Documentation Resource/Safety and execution limits ~/.config/zed/settings.json
- Windsurf AgentTwin Documentation Resource/Safety and execution limits ~/.codeium/windsurf/mcp_config.json
- Cline AgentTwin Documentation Resource/Safety and execution limits ~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json
- Gemini CLI AgentTwin Documentation Resource/Safety and execution limits ~/.gemini/settings.json
- Grok AgentTwin Documentation Resource/Safety and execution limits .mcp.json (in your project root)
- ChatGPT AgentTwin Documentation Resource/Safety and execution limits Settings → Connectors → Advanced → Developer mode
- Claude.ai AgentTwin Documentation Resource/Safety and execution limits Settings → Connectors → Add custom connector
- LangChain AgentTwin Documentation Resource/Safety and execution limits pip install langchain-mcp-adapters
Other resources on this server
- AgentTwin | Digital twins for testing AI agents
- AgentTwin enterprise pricing
- About AgentTwin
- Contact AgentTwin | AI agent testing and enterprise access
- Privacy and data handling
- AgentTwin developer resources
- AgentTwin API authentication
- AgentTwin API versioning and deprecation policy
- AgentTwin sandbox: test a synthetic workflow
- Documentation for coding agents
- AgentTwin REST API reference
- CLI and coding-agent reference
- Worlds, twins, and replicas
- Dashboard guide
- Read evidence and approve a fix
- Try the sample test
- FAQ
- Fidelity contract
- Start with your project
- Runtime requirements and verification
- AgentTwin MCP servers and connection guide
- Prepare your actual application
- Explore the browser preview
- Pull request checks
- Set up your project
- Product roadmap
- Design reproducible scenarios
- From discovery to a verified fix
- What works today
- Troubleshoot setup and tests
- Understand a timeout after a write
- How a workflow becomes a test
- Reproduce, fix, and retest
- Supported provider operations
Read from the server itself, by connecting to it and calling resources/list on 7 November 2026.
A listing does not declare its resources, so asking is the only way to know them, and this
is what the server actually offers rather than what its listing claims. Names only are
stored; connect the server for each resource's type and contents.