From e220107947074fbdeeb5b444009f9799cc4c99db Mon Sep 17 00:00:00 2001 From: ikrispin Date: Mon, 11 May 2026 14:35:56 +0300 Subject: [PATCH] fix(rh-developer): correct stale parameter names and remove redundant verification step in incident-triage Co-authored-by: Cursor --- rh-developer/skills/incident-triage/SKILL.md | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/rh-developer/skills/incident-triage/SKILL.md b/rh-developer/skills/incident-triage/SKILL.md index cd7fc0a..7dbeb5a 100644 --- a/rh-developer/skills/incident-triage/SKILL.md +++ b/rh-developer/skills/incident-triage/SKILL.md @@ -53,11 +53,10 @@ Structured incident investigation for OpenShift — traces from symptoms to root - `alertmanager_alerts` (from openshift, observability toolset) — Retrieve active Alertmanager alerts **Verification Steps:** -1. Check `openshift` server is configured in `mcps.json` -2. Check `openshift` server is configured in `mcps.json` with `observability` in its `--toolsets` -3. Verify user is logged into an OpenShift cluster (`oc whoami` succeeds) -4. Verify user has access to the target namespace(s) -5. If missing → Human Notification Protocol +1. Check `openshift` server is configured in `mcps.json` with `observability` in its `--toolsets` +2. Verify user is logged into an OpenShift cluster (`oc whoami` succeeds) +3. Verify user has access to the target namespace(s) +4. If missing → Human Notification Protocol **Human Notification Protocol:** @@ -244,12 +243,12 @@ Apply these investigation guardrails before reaching any conclusion: **MCP Tool**: `prometheus_query` (from openshift, observability toolset) **Parameters**: -- `match`: "{__name__=~\".*.*\"}" (filter by relevant metric patterns, e.g., memory, disk, connections) +- `query`: "{__name__=~\".*.*\"}" (discover available metrics by pattern, e.g., memory, disk, connections) **MCP Tool**: `prometheus_query` (from openshift, observability toolset) **Parameters**: -- `metric`: "" (confirm type before querying) +- `query`: "" (confirm metric exists and inspect its current value) **MCP Tool**: `prometheus_query` (from openshift, observability toolset)