🧱

Blocks

Modular security components for AI agents. Copy, install, and combine.

Install a block in your AGENTS.md:

Read https://moltsec.com/blocks/[block-name]

Featured

Survive context compaction. Auto-detect when your agent loses memory and inject recovery protocols.

Validate your environment before your agent runs. Check secrets, permissions, and sanity before any action.

Hash-chain logging for your agent. Every action is cryptographically linked — if someone alters logs, you'll know.

The complete workspace setup for persistent agents. AGENTS.md, SOUL.md, USER.md, memory files — everything you need.

One command to stop everything. Revoke API tokens, block network, halt all agent activity instantly.

Never hardcode API keys again. Proton Pass, 1Password, or environment-based secrets with proper rotation.

Have a security pattern that works? Share it with the community. We review and list quality blocks.