AB
AB
AiBoss

All Categories

project

IronClaw - An open-source local security AI assistant from the NearAI team

IronClaw is an open-source AI assistant developed by the NearAI team. Implemented in Rust, it prioritizes local compatibility and security. IronClaw uses a WASM sandbox for execution and manages credentials through an encrypted vault...

What is IronClaw?

IronClaw is an open-source AI assistant developed by the NearAI team. Implemented in Rust, it prioritizes local compatibility and security. IronClaw uses a WASM sandbox to isolate execution and manages credentials through an encrypted vault, ensuring sensitive data is never exposed to the LLM (Local Modeling Service). The tool supports multiple models (OpenAI, local LLM, etc.), persistent memory, a web interface, and a Docker sandbox, making it suitable for privacy-conscious individuals and teams. It also offers a one-click cloud deployment solution via NEAR AI Cloud.

IronClaw's main functions

  • Security Architecture
    • WASM SandboxSupports isolating untrusted tools to run in a WebAssembly container with the necessary permissions.
    • Encrypted safe: Ensure that credentials are injected at the host boundary and that the original key is never exposed to the LLM.
    • Injection protection prompt: Defend against attacks through pattern detection, content sanitization, and policy rules.
    • Endpoint whitelistStrictly restrict tools to accessing only pre-approved HTTP hosts and paths.
  • Intelligent Interaction
    • Multi-channel supportIt covers REPL, HTTP Webhook, Telegram/Slack, and real-time web gateways.
    • Persistent memoryImplementing a hybrid full-text and vector search using PostgreSQL and pgvector.
    • Dynamic tool buildingIt supports users in generating custom WASM tools using natural language descriptions.
    • MCP Protocol IntegrationIt has the ability to connect to external Model Context Protocol (MCP) servers for expansion.
  • Automated deployment
    • Scheduled Task EngineSupports Cron scheduling, event triggering, and Webhook for background automation.
    • Docker SandboxProvides an isolated container environment for secure execution of AI-generated code.
    • Multi-model compatibilitySupports OpenAI, native LLM, and various OpenAI-compatible endpoints.
    • Flexible deployment methodsThis includes single binary files, Docker, package managers, and cloud-based TEE environments.

How to use IronClaw

  • Installation and Deployment
    • Quick installationInstall via various methods including Windows installer, PowerShell script, Shell script, Homebrew, or Cargo compilation.
    • Database preparationCreate a PostgreSQL database and enable the pgvector extension as the data storage backend.
    • Initialization configuration:run ironclaw onboard The command launch wizard completes the database connection, NEAR AI authentication, and key encryption settings.
  • Daily use
    • Interactive Dialogue:implement cargo run or ironclaw Start REPL for command-line interaction.
    • Web interface accessUse a browser to chat, manage memories, and view task logs via the built-in web gateway.
    • Backend automationConfigure Cron schedules, event triggers, or webhooks to implement scheduled tasks and automatic responses.
  • Feature Expansion
    • Accessing external modelsIn the wizard, select an OpenAI-compatible endpoint, or set environment variables to connect to providers such as OpenRouter and Ollam.
    • Install custom toolsInstall WASM tools from the registry or describe your needs to allow the system to automatically build new tools.
    • Deployment Channel IntegrationConfigure WASM channels such as Telegram and Slack to enable multi-platform message access.
  • Operation and maintenance debugging
    • Log View:set up RUST_LOG=ironclaw=debug Run environment variables to obtain detailed debugging information.
    • Test and verification:use cargo test Run unit tests or execute specific test cases to verify functionality.

IronClaw project address

  • Project official websitehttps://www.ironclaw.com/
  • GitHub repositoryhttps://github.com/nearai/ironclaw

IronClaw Application Scenarios

  • Privacy-first personal assistantSecurely manage schedules, notes, and sensitive conversations on your local device, ensuring your data never leaves your control.
  • Enterprise security operations and maintenanceManage server credentials through an encrypted vault, automate deployment and monitoring tasks, and prevent key leaks.
  • Code security reviewRun AI-generated code analysis tools in a WASM sandbox to isolate potential malicious code and protect the development environment.
  • Multi-channel intelligent customer serviceIt integrates with Telegram, Slack, and a web interface, providing personalized auto-response services based on long-term memory.
  • Sensitive data processingLocal analysis of medical records, financial data, or legal documents, while leveraging AI capabilities to meet compliance requirements such as HIPAA.
Back