runcloud（runcloud 搜索）

<!-- version: 1.0 | updated: 2026-04-09 -->

Runcloud Skill

Query and monitor Runcloud-managed servers via the Runcloud API v3.

Setup

1. Open Workspace → Settings → API Management in the Runcloud dashboard
2. Generate an API token
3. Set environment variables:

   export RUNCLOUD_API_TOKEN="your-api-token"
   export RC="https://manage.runcloud.io/api/v3"
   export AUTH="Authorization: Bearer $RUNCLOUD_API_TOKEN"

All requests go to $RC and include the bearer token header $AUTH.

Usage

All commands use curl to hit the Runcloud API v3 and pipe through jq for readable output. Replace {serverId}, {webappId}, {databaseId}, and {cronjobId} with real IDs (fetch them from the corresponding list endpoint).

Ping (auth check)

curl -s -H "$AUTH" "$RC/ping" | jq

List servers

curl -s -H "$AUTH" "$RC/servers?page=1&perPage=40" | jq '.data[] | {id, name, ipAddress, provider, online}'

Get server

curl -s -H "$AUTH" "$RC/servers/{serverId}" | jq

Get server stats

curl -s -H "$AUTH" "$RC/servers/{serverId}/stats" | jq
# -> { "stats": { "webApplication": n, "database": n, "cronJob": n, "supervisor": n }, "country": "..." }

Get hardware info

curl -s -H "$AUTH" "$RC/servers/{serverId}/hardwareinfo" | jq
# -> kernel, processor, cpu cores, memory, disk, load average, uptime

Get latest server health

curl -s -H "$AUTH" "$RC/servers/{serverId}/health/latest" | jq
# -> { totalMemory, availableMemory, usedMemory, totalDiskSpace, availableDiskSpace, usedDiskSpace, loadAverage, updated_at }

List services

curl -s -H "$AUTH" "$RC/servers/{serverId}/services" | jq
# -> { beanstalk, httpd, mariadb, memcached, nginx, redis, supervisor } each with { realName, name, memory, cpu, running, version }

List web applications

curl -s -H "$AUTH" "$RC/servers/{serverId}/webapps?page=1&perPage=40" | jq '.data[] | {id, name, rootPath, publicPath, phpVersion}'

Get web application

curl -s -H "$AUTH" "$RC/servers/{serverId}/webapps/{webappId}" | jq

List databases

curl -s -H "$AUTH" "$RC/servers/{serverId}/databases?page=1&perPage=40" | jq '.data[] | {id, name, collation, created_at}'

Get database

curl -s -H "$AUTH" "$RC/servers/{serverId}/databases/{databaseId}" | jq

List database users

curl -s -H "$AUTH" "$RC/servers/{serverId}/databaseusers?page=1&perPage=40" | jq '.data[] | {id, username, created_at}'

List system users

curl -s -H "$AUTH" "$RC/servers/{serverId}/users?page=1&perPage=40" | jq '.data[] | {id, username, created_at}'

List cron jobs

curl -s -H "$AUTH" "$RC/servers/{serverId}/cronjobs?page=1&perPage=40" | jq '.data[] | {id, label, username, time, command, enabled}'

Get cron job

curl -s -H "$AUTH" "$RC/servers/{serverId}/cronjobs/{cronjobId}" | jq

List supervisor jobs

curl -s -H "$AUTH" "$RC/servers/{serverId}/supervisors?page=1&perPage=40" | jq '.data[] | {id, name, command, user, running}'

List SSL certificates

curl -s -H "$AUTH" "$RC/servers/{serverId}/webapps/{webappId}/ssl" | jq

List domains for a web app

curl -s -H "$AUTH" "$RC/servers/{serverId}/webapps/{webappId}/domains" | jq '.data[] | {id, name, type}'

List deployments

curl -s -H "$AUTH" "$RC/servers/{serverId}/webapps/{webappId}/git/deployments?page=1&perPage=40" | jq '.data[] | {id, status, commit_hash, created_at}'

Non-destructive writes (POST)

These endpoints *trigger* safe actions — they do not create billable resources or delete anything. Still, confirm with the operator before running them on a production server.

Trigger a new deployment

curl -s -X POST -H "$AUTH" -H "Content-Type: application/json" \
  "$RC/servers/{serverId}/webapps/{webappId}/git/deployments" | jq

Test a cron job (runs once now)

curl -s -X POST -H "$AUTH" -H "Content-Type: application/json" \
  "$RC/servers/{serverId}/cronjobs/{cronjobId}/test" | jq

Restart a service

curl -s -X POST -H "$AUTH" -H "Content-Type: application/json" \
  -d '{"action":"restart"}' \
  "$RC/servers/{serverId}/services/{serviceRealName}/action" | jq
# serviceRealName examples: nginx-rc, apache2-rc, mysql, redis-server, supervisor

Deploy Let's Encrypt SSL

curl -s -X POST -H "$AUTH" -H "Content-Type: application/json" \
  -d '{"provider":"letsencrypt","enableHttp":true,"enableHsts":false,"ssl_protocol_id":1}' \
  "$RC/servers/{serverId}/webapps/{webappId}/ssl" | jq

Notes

• Auth: every request needs Authorization: Bearer $RUNCLOUD_API_TOKEN. The token grants full workspace access — treat it like a password.
• Pagination: array endpoints accept ?page=N&perPage=M (max perPage=40). Responses wrap data in { "data": [...], "meta": { "pagination": {...} } }.
• Rate limiting: check the X-RateLimit-Limit and X-RateLimit-Remaining headers on every response; a 429 means back off and retry later.
• Finding IDs: serverId, webappId, databaseId, cronjobId all come from the matching list endpoint — always list first if you don't know the ID.
• v3 is still in progress: if an endpoint 404s, the v2 path at https://manage.runcloud.io/api/v2 usually mirrors it one-to-one. Docs index: https://runcloud.io/docs/api/v3/doc-625011
• Safety: this skill deliberately omits every DELETE, resource create, and password-change endpoint. Do not add them without an explicit operator request.

Examples

# Health snapshot of every server (one-liner)
curl -s -H "$AUTH" "$RC/servers?perPage=40" \
  | jq -r '.data[].id' \
  | while read id; do
      echo "=== server $id ==="
      curl -s -H "$AUTH" "$RC/servers/$id/health/latest" \
        | jq '{loadAverage, usedMemory, availableMemory, usedDiskSpace}'
    done

# Find a specific server by name
curl -s -H "$AUTH" "$RC/servers?perPage=40" \
  | jq '.data[] | select(.name | contains("prod"))'

# Flag servers whose load average is above 1.0
curl -s -H "$AUTH" "$RC/servers?perPage=40" \
  | jq -r '.data[].id' \
  | while read id; do
      load=$(curl -s -H "$AUTH" "$RC/servers/$id/health/latest" | jq '.loadAverage')
      awk -v l="$load" -v i="$id" 'BEGIN{ if (l+0 > 1.0) print "HOT:", i, l }'
    done

# Count web apps, databases, and cronjobs on a server
curl -s -H "$AUTH" "$RC/servers/{serverId}/stats" \
  | jq '.stats'