Claw-CLI v1.0.0 – Secure Agent Ready for Production & ClawCloud

Security-First Design: Unlike other agents, Claw-CLI operates on a "fail-closed" principle. Every action is explicitly allow-listed and validated, ensuring the LLM cannot execute arbitrary or unsafe commands.
Sandboxed Execution: All web automation occurs within an isolated browser environment, preventing unintended system access.
User Confirmation for Sensitive Actions: Critical operations (e.g., sending messages) require explicit user approval. You're always in control.
Auditable Logs: Every agent action and decision is logged for transparency and security auditing.
Local-First, Single-User: Designed for your personal machine, offering robust security without the complexities of multi-user environments.

v0.1.0-alpha

docs: add getting started instructions to README.md