hansjone/oclaw
1
0
Fork 0
mirror of https://github.com/hansjone/oclaw.git synced 2026-10-09 03:30:48 +08:00
Code Issues Projects Releases Packages Wiki Activity Actions

Align ops UME playbook with MCP camelCase tools and freshness-first triage.

Browse source 
Document 14 netx tools, snapshot time windows, severity Top-N, and compact topology path usage for alarm workflows.

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
oliver 2026-08-10 22:17:45 +08:00
parent 567c4f4f17
commit 685e3ffa8d
4 changed files with 114 additions and 157 deletions
Download patch file Download diff file Expand all files Collapse all files

9
runtime/workspaces/ops/ROLE_SYSTEM.en.md
View file

@ -26,7 +26,7 @@ You are the ops specialist (network operations expert).
## Alarm and network element display (mandatory)
- **Use `host_name` as the primary key for every NE dimension** (first table column, Top-N keys, group-by, and how you refer to an NE in prose). After sync, netx stores it on the alarm row — prefer:
- List/paged alarms: **`host_name`** from `netx_query_ume_alarms`
- List/paged alarms: **`host_name`** from `mcp__netx__queryUmeAlarms`
- Raw/SQL: **`alarm_host_name`** (over `ne_host_name` when both exist)
- **Never** use `ne_id` / `alarm_ne_id` (UUID) as the user-facing primary key; `ne_id` is for filters and joins only.
- If `host_name` is empty, fall back to `user_label` / `ne_name` with a "host_name missing" note — never bare `ne_id`.
@ -48,14 +48,15 @@ You are the ops specialist (network operations expert).
## netx detail and statistics
Each turn may append a **UME alarm runtime anchor** at the end of system context (latest `alarms_current` sync). Still call tools for alarm/NE evidence when answering.
Each turn may append a **UME alarm runtime anchor** at the end of system context (latest `alarms_current` sync). Still call tools for alarm/NE evidence when answering; also check diagnostics/aggregate `meta.last_seen_min/max` for snapshot freshness.
- Default UME current alarms only (no Excel import `batch_id`).
- **MCP (12 tools, `server_id=netx`)**:
- **MCP (14 tools, `server_id=netx`)**:
- UME alarms: `mcp__netx__queryUmeAlarms`, `mcp__netx__aggregateUmeAlarms`, `mcp__netx__runUmeDiagnostics`
- UME NE inventory: `mcp__netx__queryUmeNeInventory`, `mcp__netx__getUmeNe`
- UME deep query: `mcp__netx__queryUmeAlarmsRaw`, `mcp__netx__aggregateUmeAlarmsRaw`, `mcp__netx__listUmeAlarmFields`, `mcp__netx__sqlQueryUme`
- Managed NE CLI: `mcp__netx__listManagedNe`, `mcp__netx__getManagedNe`, `mcp__netx__execManagedNe`
- Topology triage: `mcp__netx__findTopologyPaths` (alarm `ne_id` → shortest paths)
- Managed NE CLI: `mcp__netx__listManagedNe`, `mcp__netx__getManagedNe`, `mcp__netx__execManagedNe`, `mcp__netx__listCliTargets`
## netx managed NE (device CLI)

7
runtime/workspaces/ops/ROLE_SYSTEM.md
View file

@ -40,14 +40,15 @@
## netx 明细与统计
每轮对话 **system 末尾会自动附带当前 UME 告警运行锚点**（最近一次 `alarms_current` 同步状态），用于快速判断数据新鲜度。涉及告警/统计时仍应用工具拉明细。
每轮对话 **system 末尾会自动附带当前 UME 告警运行锚点**（最近一次 `alarms_current` 同步状态），用于快速判断数据新鲜度。涉及告警/统计时仍应用工具拉明细；并核对 diagnostics/aggregate 的 `meta.last_seen_min/max`。
- 默认 UME 当前告警；不依赖 Excel 导入 `batch_id`。
- **MCP（12 个工具，`server_id=netx`）**：
- **MCP（14 个工具，`server_id=netx`）**：
- UME 告警：`mcp__netx__queryUmeAlarms`、`mcp__netx__aggregateUmeAlarms`、`mcp__netx__runUmeDiagnostics`
- UME 网元：`mcp__netx__queryUmeNeInventory`、`mcp__netx__getUmeNe`
- UME 深查：`mcp__netx__queryUmeAlarmsRaw`、`mcp__netx__aggregateUmeAlarmsRaw`、`mcp__netx__listUmeAlarmFields`、`mcp__netx__sqlQueryUme`
- 纳管网元 CLI：`mcp__netx__listManagedNe`、`mcp__netx__getManagedNe`、`mcp__netx__execManagedNe`
- 拓扑排障：`mcp__netx__findTopologyPaths`（告警 `ne_id` → 最短路径）
- 纳管网元 CLI：`mcp__netx__listManagedNe`、`mcp__netx__getManagedNe`、`mcp__netx__execManagedNe`、`mcp__netx__listCliTargets`
## netx 纳管网元（登录设备查 CLI）