菜单
This skill should be used when the user asks to "deploy to production", "restart a service", "do a rolling update", "canary deploy", "helm upgrade", or "rollout to K8s". Manual-only for safety — requires explicit /ssh-deploy invocation.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
This skill should be used when the user asks to "run a security scan", "check for vulnerabilities", "audit compliance", "check certificate expiry", "scan open ports", or mentions security reviews, CIS benchmarks, or compliance assessments on remote servers.
This skill should be used when the user asks to "show me the available tools", "what Docker tools exist", "search for kubernetes tools", "explore the tool catalog", or wants to browse the 476 bridge-mcp tools by group or keyword.
This skill should be used when the user asks to "run a command on a server", "check Docker containers", "restart a service", "read a remote file", "check disk usage", or mentions managing remote infrastructure via SSH. Teaches CLI-first workflow with progressive tool discovery.
Use when the user wants to run a command on or manage a remote host over SSH — a server, Raspberry Pi, VM, NAS, network switch, or air-gapped machine. Triggers: "check/manage my server", "run X on host Y", "ssh into / exec on host", "restart nginx on prod", "docker ps / containers on", "kubectl / k8s pods / helm on the cluster", "tail the logs on", "disk usage / cpu / memory / processes on host", "what's listening on port", "is host reachable", "service status", "configure / validate the bridge". Drives the bridge-mcp CLI (476 SSH tools) — guided setup, status checks, and tool execution.
Use when the user wants to explore, search, or learn what bridge-mcp can do: "what tools are available", "list bridge tools", "find a tool for X", "show docker / kubernetes / systemd / network tools", "search tools for logs / firewall / packages / databases", "which tool groups exist", "what can the bridge do". Progressive discovery across the 476 tools and 77 groups, with token-efficient describe-tool output.
Use when the user says "bridge", "remote", "ssh tool", "run on host", "check host", "configure host", "setup bridge", or wants to manage remote servers or bridge config via CLI.
| name | ssh-deploy |
| description | This skill should be used when the user asks to "deploy to production", "restart a service", "do a rolling update", "canary deploy", "helm upgrade", or "rollout to K8s". Manual-only for safety — requires explicit /ssh-deploy invocation. |
| user-invocable | true |
| disable-model-invocation | true |
| argument-hint | <host> <action> |
| compatibility | 2.1+ |
Deploy changes to remote infrastructure via bridge-mcp CLI.
This skill is manual-only — invoke with /ssh-deploy.
Parse $ARGUMENTS: first word = host, second word = action (restart, canary, rolling, k8s, helm).
Before ANY deployment, verify:
Health check — confirm the target is healthy BEFORE changing anything:
bridge-mcp tool ssh_diagnose host=HOST --json
Confirm with user — always ask before proceeding with destructive operations.
Know the rollback — have a rollback command ready before executing.
# Check current status
bridge-mcp tool ssh_service_status host=HOST service=SERVICE --json
# Restart
bridge-mcp tool ssh_service_restart host=HOST service=SERVICE
# Verify recovery
bridge-mcp tool ssh_service_status host=HOST service=SERVICE --json
Deploy to a single canary host first, verify, then decide:
# Execute on canary
bridge-mcp tool ssh_canary_exec host=HOST command="DEPLOY_COMMAND"
# Health check canary
bridge-mcp tool ssh_diagnose host=HOST --json
# ASK the user before continuing to remaining hosts
Execute on one host at a time with health checks between:
bridge-mcp tool ssh_rolling_exec host=HOST command="DEPLOY_COMMAND"
# Check current state
bridge-mcp tool ssh_k8s_get host=HOST resource=deployments namespace=NAMESPACE --json
# Apply update
bridge-mcp tool ssh_k8s_apply host=HOST manifest="PATH_TO_YAML" namespace=NAMESPACE
# Watch rollout
bridge-mcp tool ssh_k8s_rollout host=HOST action=status deployment=DEPLOYMENT namespace=NAMESPACE
# Rollback if needed
bridge-mcp tool ssh_k8s_rollout host=HOST action=undo deployment=DEPLOYMENT namespace=NAMESPACE
# Check current release
bridge-mcp tool ssh_helm_status host=HOST release=RELEASE --json
# Upgrade
bridge-mcp tool ssh_helm_upgrade host=HOST release=RELEASE chart=CHART
# Verify
bridge-mcp tool ssh_helm_status host=HOST release=RELEASE --json
# Rollback if needed
bridge-mcp tool ssh_helm_rollback host=HOST release=RELEASE
After every deployment:
Health check — verify the system is healthy:
bridge-mcp tool ssh_diagnose host=HOST --json
Verify service — check the specific service/app is running:
bridge-mcp tool ssh_service_status host=HOST service=SERVICE --json
Report — tell the user: what was deployed, current status, rollback command if needed.