Security Best Practices
MemantoClaw enforces security at four layers: Network, Filesystem, Process, and Inference.Layer Protections
- Network Layer: Deny-by-default egress. Configured via OpenShell policy. Binary-Scoped rules ensure only authorized binaries (like
ghorgit) can access specific endpoints. - Filesystem Layer: Uses Landlock LSM + container mounts.
/sandboxis read-only, while specific paths like/sandbox/.openclaw-dataand/tmpare writable. Gateway config (/sandbox/.openclaw) is immutable and hash-pinned. - Process Layer: Drops dangerous Linux capabilities using
capsh. Setsulimit -u 512to mitigate fork-bomb attacks. Enforcesno-new-privilegesto block privilege escalation via setuid binaries. Removes build toolchains (gcc,make) andnetcatfrom the image. - Inference Layer: Routes model API calls to controlled backends via
inference.local.
Credential Storage
Credentials (likeMOORCHEH_API_KEY, OPENAI_API_KEY) are stored in plaintext JSON at:
~/.memantoclaw/credentials.json
They are created with mode 0600 on the host. They are never injected into the sandbox. The host bridge authenticates requests before forwarding them. If you suspect exposure, rotate keys and remove the stored file:
OpenClaw Controls
MemantoClaw delegates application-layer security to OpenClaw. OpenClaw provides:- Prompt Injection Detection: Neutralizes attempts like
<system>tag spoofing. - Tool Access Control: High-risk tools (
exec,spawn,fs_write) are gated by a multi-layer policy pipeline. - Environment Variable Security: Blocks dangerous env vars (
NODE_OPTIONS,LD_PRELOAD). - Secret Scanner: Intercepts writes targeting memory paths that look like API keys before they reach the disk.
For complete, unabridged technical details on this topic, refer to the official NVIDIA NemoClaw Documentation. Portions of this guide are summarized and adapted from NVIDIA Corporation (Copyright © 2026), licensed under the Apache License, Version 2.0.