OpenClaw and personal-agent gateways on local models
Parent: Local LLM Inference on Consumer Hardware (RTX 5080 eGPU, Linux, Remote Access) · Topic entry · 7 branches · skill ai-agents-orchestration
A published reference is not available for this topic yet.
Children
- OpenClaw history and renames (Warelay, Clawdbot, Moltbot, foundation) (frontier)
- OpenClaw Gateway architecture (daemon, channels, skills, memory, heartbeat, sandbox) (frontier)
- OpenClaw local model providers (Ollama native API, LM Studio Responses API, OpenAI-compatible servers) (frontier)
- Local model capability floors (64k context, tool-calling reliability, injection resistance) (frontier)
- OpenClaw security record (exposed gateways, CVE-2026-25253, ClawHavoc supply chain) (frontier)
- OpenClaw hardening baseline (loopback bind, token auth, DM pairing, tool deny, sandbox) (frontier)
- Personal-agent gateway alternatives (Open WebUI, LibreChat, Hermes Agent, goose, n8n, nanobot, IronClaw) (frontier)
Frontier under this node: Local model capability floors (64k context, tool-calling reliability, injection resistance), OpenClaw Gateway architecture (daemon, channels, skills, memory, heartbeat, sandbox), OpenClaw hardening baseline (loopback bind, token auth, DM pairing, tool deny, sandbox), OpenClaw history and renames (Warelay, Clawdbot, Moltbot, foundation), OpenClaw local model providers (Ollama native API, LM Studio Responses API, OpenAI-compatible servers), OpenClaw security record (exposed gateways, CVE-2026-25253, ClawHavoc supply chain), Personal-agent gateway alternatives (Open WebUI, LibreChat, Hermes Agent, goose, n8n, nanobot, IronClaw)