Forum ForumsWhat’s NewRecent PostsMembers Forums Search Notifications Clear all Tag: local_llm Search Phrase: Search Type: Search Entire Posts Search Titles Only Find Topics by Tags Find Posts by User Find Topics Started by User Advanced search options Search in Forums: OpenClaw — Architecture and Threat Modeling — — Trust Boundaries and Component Isolation — — Attack Surface Mapping — — Threat Model Templates and Examples — Sandboxing and Execution Isolation — — Container and Runtime Hardening — — Sandbox Escapes and Breakout Research — — Seccomp, AppArmor, and LSM Profiles — Credential and Secret Handling — — Secret Injection Patterns — — Credential Leakage via Agents and Logs — Network Egress and Exfiltration Controls — — Egress Filtering Configurations — — Detecting Agent Exfiltration Attempts — Plugin and Tool Security — — Tool Vetting and Review — — MCP and Tool Protocol Security — — Supply Chain Integrity for Tools The Claw Family — NemoClaw — NVIDIA Privacy and Security Stack — — GPU Memory Isolation and Leakage — — NIM Container Security — — NeMo Guardrails — Security vs. Privacy Tradeoffs — NanoClaw — Container-Isolated Anthropic Agent SDK — — Container Isolation Model and Gaps — — Anthropic Agent SDK Security Surface — — Hardening NanoClaw Deployments — IronClaw — NEAR AI Encrypted Enclave Runtime — — Enclave Attestation and Verification — — Side Channel Risks in Enclave Deployments — — Key Management and Sealed Storage — — NEAR AI Integration Security — Comparing Claw Family Runtimes Non-Claw Alternatives — Coding Agents — Claude Code, Cursor, Aider, OpenHands — — Claude Code Security — — Cursor Security — — Aider and OpenHands Security — Browser and Operator Agents — OpenAI Operator, Goose — — OpenAI Operator Security — — Goose (Block) Security — Code-First Agent Frameworks — LangGraph, CrewAI, AutoGen, SuperAGI — — LangGraph Security — — CrewAI and AutoGen Security — — SuperAGI Security — Cross-Framework Security Comparisons Security Patterns and Hardening — Prompt Injection Defenses — — Indirect Injection via Tools and Retrieved Data — — Injection Detection and Runtime Monitoring — — Benchmarks and Evaluation Methodologies — Sandboxing Strategies for Agent Runtimes — — MicroVMs and gVisor for Agent Isolation — — WebAssembly as an Agent Sandbox — — Default Sandbox Configurations Are Insufficient — Credential and Secret Management Patterns — — Vault Integration Patterns — — Scoped and Ephemeral Credentials for Agents — Network Egress Controls — — Allowlist Design for Agent Network Access — — DNS and Layer 7 Egress Controls — Supply Chain Integrity for Agent Runtimes — — SBOM Generation and Artifact Signing — — Dependency Auditing and Pinning Enterprise and Regulated Deployments — Compliance Framework Mapping — — SOC 2 and ISO 27001 for Agent Runtimes — — HIPAA and Healthcare Agent Deployments — — FedRAMP and Government Deployments — Audit Logging and Security Observability — — Agent Audit Log Design — — SIEM Integration for Agent Events — Enclave Deployments and Confidential Computing — — TEE Platform Comparison for Agent Workloads — — Operational Security for Enclave Deployments — CISO Evaluation Guides — — Vendor Security Questionnaires — — Self-Hosted vs. Vendor-Hosted Risk Tradeoffs Community — Announcements — Introductions — Show and Tell — News and Vulnerability Disclosures — Off-Topic Main Category — Main Forum Search in date period: Any Date Last 24 hours Last Week Last Month Last 3 Months Last 6 Months Last Year ago Sort Search Results by: Relevancy Date User Forum Descending order Ascending order Page 2 / 4 Prev 1 2 3 4 Next Kubernetes NetworkPolicies vs service mesh for agent networking - which is simpler? Ella Morozov 1 month ago self-hosting local llms agent frameworks nano claw ironclaw Walkthrough: Implementing a mandatory approval step for all agent tool calls. Mia C. 1 month ago raspberry pi linux admin open claw ironclaw local llm Thoughts on using Aider only in a read-only file system snapshot? Sam Ortega 1 month ago self-hosting docker-compose local-llm openclaw nano-claw Direct file system access vs MCP file server - which is less risky? Sam Ortega 1 month ago self-hosting docker-compose local-llm openclaw nano-claw Hot take: 'Tool call' and 'tool result' logs need separate verbosity controls. Mia C. 1 month ago raspberry pi linux admin open claw ironclaw local llm How do you configure IronClaw to minimize Spectre v2 exposure? Hal Newb 2 months ago linux_basics openclaw local_llm home_lab Beginner question: Does MCP have any built-in encryption, or is it all on me? Lars J. 2 months ago selfhosting docker_compose local_llm_deployment ironclaw nanoClaw Opinion: The documentation's 'quick start' should include security flags from day one. Marta Reyes 2 months ago docker selfhosting local_llms openclaw_deployment agent_containers Is it safe to block all *.amazonaws.com except my S3 model bucket? Hal Newb 2 months ago linux_basics openclaw local_llm home_lab What's the current best practice for mounting models read-only? Marta Reyes 2 months ago docker selfhosting local_llms openclaw_deployment agent_containers Vault for secrets vs environment variables - which is less likely to leak via an agent? Mia C. 2 months ago raspberry pi linux admin open claw ironclaw local llm Opinion: The 'session_id' field is the most critical one for any correlation. Luke M. 2 months ago local llm inference security model isolation openclaw nano claw Just built a stress test that spawns 500 containers to find the orchestrator limit Luke M. 2 months ago local llm inference security model isolation openclaw nano claw Unpopular opinion: Running any third-party "skills" from the community hub is asking for trouble. Mia C. 2 months ago raspberry pi linux admin open claw ironclaw local llm Step-by-step: Configuring OpenClaw to log to a remote syslog server with TLS. Sam Ortega 2 months ago self-hosting docker-compose local-llm openclaw nano-claw Page 2 / 4 Prev 1 2 3 4 Next Share: Share Tweet Share