H3C iMC REST API MCP wrapper for network device management, alarms, performance monitoring, and topology queries. Provides read-only access to iMC v7.3 REST endpoints with HTTP Digest authentication.
Strong foundation with 15 well-named tools following verb_noun convention (list_*, get_*). All tools have descriptions (avg ~180 chars, within 10-1024 baseline). Input schemas are present with type declarations and parameter descriptions. However, output schemas are undocumented, responses are raw API JSON without field documentation. Error handling is basic (generic error messages without recovery guidance). No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite all being READ-ONLY. Parameter validation is solid (numeric ID checks, safe param checks) but not exposed in descriptions. Missing: enum constraints for status/level fields, explicit pagination guidance in descriptions, and structured output documentation.
查询单条告警的完整详情(告警描述、根因摘要 reason、恢复建议 suggestion、关联设备、发生/恢复/确认时间、定位信息等)。alarmId 来自 imc_list_alarms。在排查设备离线或故障时用于获取告警的完整上下文。iMC REST: GET /imcrs/fault/alarm/{alarmId}
查询指定时间范围内的告警统计(按时间分组的告警数量、级别分布等)。需传 Unix 毫秒时间戳。用于生成告警趋势报告。iMC REST: GET /imcrs/fault/statistics
查询单个设备的详细信息(基本属性、型号、固件版本、运行状态、运行时长、SysDescription 等)。deviceId 来自 imc_list_devices。iMC REST: GET /imcrs/plat/res/device/{deviceId}
查询指定设备已配置的性能监控实例列表(返回 taskId/instanceId,用于后续 imc_get_perf_monitor 查询)。iMC REST: GET /imcrs/perf/monitor/device?devId=X
查询单条链路的详细信息(源/目标设备、接口描述、带宽、状态等)。需 linkId+topoId(linkId 来自 imc_list_links)。iMC REST: GET /imcrs/plat/res/link/{linkId}?topoId=X
查询指定设备+指标的性能监控详情(CPU/内存/接口流量等实时数据)。需 devId+taskId(taskId 来自 imc_get_device_perf)。iMC REST: GET /imcrs/perf/monitor?devId=X&taskId=Y
Output schemas undocumented. Tools return raw API JSON (prettyJSON wrapper) without field documentation. LLMs cannot plan downstream calls or extract specific fields reliably.
No tool annotations despite all 15 tools being READ-ONLY. Missing readOnlyHint declarations that would signal to agents these are safe to retry and have no side effects.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 66 | 2026-07-28+ | v2 |
查询最近 N 条实时告警(无需过滤条件,按发生时间倒序)。operatorName 自动注入。用于快速了解当前网络告警态势。iMC REST: GET /imcrs/fault/faultRealTime
查询网络拓扑(节点+链路)。topoId 缺省 1=全网拓扑。返回设备节点坐标、运行状态、链路连接关系等。iMC REST: GET /imcrs/plat/res/topo/topoNodeAndLink/{topoId}
查询 iMC 告警列表。支持按告警级别、恢复/确认状态、设备、时间范围等过滤。operatorName 自动使用当前 MCP 连接的操作员账号。返回告警 ID、设备信息、告警级别、描述、发生/恢复时间、确认状态等。iMC REST: GET /imcrs/fault/alarm
查询 iMC 中配置的设备分组列表(分组 ID、名称、描述、包含设备等)。用于了解设备的逻辑组织方式,按分组维度统计设备健康度。iMC REST: GET /imcrs/plat/device/group
列出指定设备的网络接口(接口名、类型、管理状态、运行状态、速率、IP 等)。deviceId 来自 imc_list_devices。iMC REST: GET /imcrs/plat/res/device/{deviceId}/interface
查询 iMC 纳管的网络设备列表。返回设备概要信息(deviceId、设备标签/名称、管理 IP、状态、型号、MAC 地址等)。这是获取 deviceId 的入口,设备详情/接口/性能查询、告警按设备过滤都依赖该 ID。iMC REST: GET /imcrs/plat/res/device
查询指定拓扑的链路列表(链路 ID、源/目标设备、接口、状态等)。需 topoId(来自 imc_get_topology)。iMC REST: GET /imcrs/plat/res/link?topoId=X
查询全局性能监控指标列表(CPU/内存/接口流量等,含阈值设置、采集间隔等)。iMC REST: GET /imcrs/perf/task
查询根源告警列表(已去重的根本原因告警,用于快速定位故障源头)。iMC REST: GET /imcrs/fault/rootAlarm
Error messages are generic ('查询设备列表失败', '查询设备详情失败'). No recovery guidance or actionable next steps. LLMs cannot self-correct or determine if errors are retryable.
Enum constraints missing for status/level fields. 'alarmLevel' accepts '1,2,3,4' but not declared as enum. 'status' field in imc_list_devices lacks valid values. LLMs may hallucinate invalid values.
Pagination guidance incomplete. Tools accept 'start' and 'size' but descriptions do not state max limits, default behavior, or total count in response. Agents cannot reliably iterate through large result sets.