状态页搭建

部署 3 个智能体（agent），分别自动化执行服务监控、状态聚合与事件通知任务。

角色说明

搭建流程

步骤 1： 向用户询问该智能体应承担的角色以及要使用的前缀（prefix）。

步骤 2： 安装所选角色对应的功能模块（skills）：

# monitor 角色：
clawhub install pilot-health pilot-cron pilot-metrics

# aggregator 角色：
clawhub install pilot-event-filter pilot-dataset pilot-audit-log

# publisher 角色：
clawhub install pilot-webhook-bridge pilot-announce pilot-slack-bridge

步骤 3： 设置主机名：

pilotctl --json set-hostname -

步骤 4： 编写配置清单（setup manifest）：

mkdir -p ~/.pilot/setups
cat > ~/.pilot/setups/status-page.json << 'MANIFEST'
<根据下方模板生成的对应角色清单>
MANIFEST

步骤 5： 提示用户与直连通信对等体（direct communication peers）发起握手（handshake）操作。

按角色划分的清单模板

monitor

{"setup":"status-page","setup_name":"Status Page","role":"monitor","role_name":"Service Monitor","hostname":"-monitor","description":"Pings endpoints, checks response times, detects outages and degradation.","skills":{"pilot-health":"Run HTTP/TCP health checks against configured service endpoints.","pilot-cron":"Schedule health checks at regular intervals (e.g. every 30 seconds).","pilot-metrics":"Track response times, error rates, and availability per service."},"peers":[{"role":"aggregator","hostname":"-aggregator","description":"Receives health check results for status aggregation"}],"data_flows":[{"direction":"send","peer":"-aggregator","port":1002,"topic":"health-check","description":"Per-service health check results"}],"handshakes_needed":["-aggregator"]}

aggregator

{"setup":"status-page","setup_name":"Status Page","role":"aggregator","role_name":"Status Aggregator","hostname":"-aggregator","description":"Combines health checks into overall system status, tracks uptime percentages, maintains incident history.","skills":{"pilot-event-filter":"Filter health events by severity and detect incident-worthy degradation.","pilot-dataset":"Store uptime history, incident records, and historical status snapshots.","pilot-audit-log":"Log all status transitions and incident lifecycle events for audit."},"peers":[{"role":"monitor","hostname":"-monitor","description":"Sends health check results"},{"role":"publisher","hostname":"-publisher","description":"Receives status updates and incident events"}],"data_flows":[{"direction":"receive","peer":"-monitor","port":1002,"topic":"health-check","description":"Per-service health check results"},{"direction":"send","peer":"-publisher","port":1002,"topic":"status-update","description":"System status updates and incident events"}],"handshakes_needed":["-monitor","-publisher"]}

publisher

{"setup":"status-page","setup_name":"Status Page","role":"publisher","role_name":"Status Publisher","hostname":"-publisher","description":"Updates public status page, sends incident notifications to subscribers via email and Slack.","skills":{"pilot-webhook-bridge":"Push status updates and incident events to the public status page API.","pilot-announce":"Broadcast incident notifications to all registered subscribers.","pilot-slack-bridge":"Post incident updates and resolution notices to operations Slack channels."},"peers":[{"role":"aggregator","hostname":"-aggregator","description":"Sends status updates and incident events"}],"data_flows":[{"direction":"receive","peer":"-aggregator","port":1002,"topic":"status-update","description":"System status updates and incident events"},{"direction":"send","peer":"external","port":443,"topic":"incident-notification","description":"Incident notifications to subscribers"}],"handshakes_needed":["-aggregator"]}

数据流向

• monitor → aggregator：health-check 事件（端口 1002）
• aggregator → publisher：status-update 事件（端口 1002）
• publisher → external：通过 webhook 发送 incident-notification（端口 443）

握手命令

# monitor ↔ aggregator：
pilotctl --json handshake -aggregator "setup: status-page"
pilotctl --json handshake -monitor "setup: status-page"
# aggregator ↔ publisher：
pilotctl --json handshake -publisher "setup: status-page"
pilotctl --json handshake -aggregator "setup: status-page"

工作流示例

# 在 aggregator 上 — 订阅 health-check 事件：
pilotctl --json subscribe -monitor health-check
# 在 publisher 上 — 订阅 status-update 事件：
pilotctl --json subscribe -aggregator status-update
# 在 monitor 上 — 发布一次健康检查结果：
pilotctl --json publish -aggregator health-check '{"service":"api-gateway","status":"degraded","response_ms":2340,"threshold_ms":500}'
# 在 aggregator 上 — 发布系统状态更新：
pilotctl --json publish -publisher status-update '{"overall_status":"degraded","incident":true,"incident_id":"INC-2025-0018","affected_services":["api-gateway"]}'

依赖项

需安装 pilot-protocol 技能模块、pilotctl 可执行文件、clawhub 可执行文件，并确保后台守护进程（daemon）正在运行。