NanoClaw NEW

Nanoclaw (gavrielc/nanoclaw) is an open-source personal Claude assistant that runs the Anthropic Agents SDK and connects to WhatsApp.It runs agents in containerized environments (Apple containers on macOS and Linux containers) with filesystem isolation and permission checks.

Features include memory, scheduled jobs, skills, and multiple channel providers for persistent context and automated workflows.A single-process, compact codebase simplifies customization, debugging, and on-premise deployment for developers and small teams.

Setup tooling includes an installation wizard, setup scripts, a monitoring dashboard, and debugging utilities to streamline operations.Application-level security controls—allowlists, pairing codes, and per-group session isolation—help limit access and reduce risk.

Hosted on GitHub under the MIT license, the repository provides configuration examples, skill extensions, and integration points for CI/CD and developer workflows.

Key Features

• Anthropic Agents SDK runtime with WhatsApp integration
• Containerized agent execution (Apple containers on macOS and Linux) with filesystem isolation and permission checks
• Built-in memory, scheduled jobs, and extensible skills for automated workflows
• Multiple channel providers for persistent context and multi-channel integration
• Application-level security controls (allowlists, pairing codes, per-group session isolation)

Use Cases

• Use nanoclaw as a secure on-prem personal assistant that runs containerized, filesystem-isolated Claude agents to handle sensitive customer or HR queries over WhatsApp while enforcing session isolation, memory controls and enterprise security policies
• Automate IT and business workflows by scheduling containerized agents to run backups, generate compliance or analytics reports, and push incident alerts or summaries to teams via WhatsApp, with centralized monitoring on the dashboard
• Deploy a developer and product-support assistant on-prem that leverages multi-channel context, persistent memory and custom skills to answer technical questions, produce code snippets or documentation, and notify stakeholders privately over WhatsApp while keeping all data local

Who is it for?

• Software developers
• Engineering teams
• Devops engineers
• Machine learning engineers
• Open source contributors