videomemory（videomemory 运维）

VideoMemory Setup

Use this skill when the user wants OpenClaw to install, start, relaunch, or check VideoMemory.

For requests like "install videomemory please", "set up videomemory", or "Install the VideoMemory skill from ClawHub and send me the UI", prefer installing the OpenClaw plugin package. If the plugin install path is unavailable, run the safe onboarding command and reply with the returned UI link.

Preferred Commands

First-class plugin install:

openclaw plugins install @clamepending/videomemory@0.1.7

After the plugin is installed and the gateway has restarted, prefer the plugin tools:

• videomemory_onboard
• videomemory_relaunch
• videomemory_status

CLI fallback from the OpenClaw workspace root after clawhub install videomemory, inspect the onboarding plan:

npx -y @clamepending/videomemory@0.1.7 onboard --safe --repo-ref v0.1.3 --explain

If the plan matches the user's request, run onboarding:

npx -y @clamepending/videomemory@0.1.7 onboard --safe --repo-ref v0.1.3

Relaunch:

npx -y @clamepending/videomemory@0.1.7 relaunch --repo-ref v0.1.3

Check status:

curl -fsS http://127.0.0.1:5050/api/health

What Onboarding Does

• Starts VideoMemory without Docker.
• Installs the OpenClaw bridge files used by VideoMemory.
• Returns the user-facing VideoMemory UI link.
• Runs in safe mode, avoiding network-exposure setup, chat notifications, and privileged setup paths.

Ground Rules

• Prefer the published VideoMemory CLI command above over hand-written setup commands.
• Prefer openclaw plugins install @clamepending/videomemory@0.1.7 when OpenClaw plugin installation is allowed.
• Run --safe --explain before onboarding when acting from chat.
• If onboarding or relaunch fails, report the actual stderr instead of guessing.
• After a successful onboarding or relaunch, reply with the returned UI link.