Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →mohitagw15856/escalation-tree
Design a support/incident escalation tree — who handles what, when it escalates, and to whom.
Passed 19 of 22 test cases with the skill, and 18 of 22 without it.
3 fixed16 passed either way1 still failing2 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 3 test cases the skill fixed. Same model, same prompt, with and without the skill.
Our principal engineers complain that every customer inquiry is currently paged directly to them. Build an escalation structure that prevents this.
| Checked | Without skill | With skill |
|---|---|---|
| Initial ticket triage and runbook resolutions are constrained to Tier 1 before reaching senior engineering | Missed | Passed |
The instructions your model reads when the skill loads.
Escalation goes wrong two ways: things sit too long before someone senior is pulled in, or everything gets escalated and senior people drown. A clear escalation tree fixes both — it defines the tiers, the severity that sets the path, the time triggers that force escalation, and who owns each step. This skill designs that, so the right person is on the right issue at the right time.
Ask for these only if they aren't already provided:
1. Severity levels — define each (SEV1/P1 … or Critical/High/Normal/Low) with concrete criteria — what qualifies, blast radius, and the response & resolution targets per level. Ambiguous severity is why escalation fails.
2. The tiers — who owns what:
| Tier | Owns | Can resolve | Escalates when |
|---|---|---|---|
| Tier 1 | first response, known issues | runbook items | unresolved in [time] or sev ≥ [x] |
| Tier 2 | deeper diagnosis | most issues | needs code/infra change |
| Eng on-call | code/infra | the system | — |
3. The tree (routing) — by severity, the path and the time triggers:
SEV1 → page eng on-call immediately + notify manager; if unacked in 5 min → secondary; if 15 min → eng lead. Normal → tier-1; if unresolved in 1 business day → tier-2.
Show the branch logic clearly (who, after how long, to whom).
4. Contacts & roles — by role (not just names — names change): who fills each, primary/secondary, and how they're reached per severity (page vs. Slack vs. ticket).
5. Customer communication — the update cadence per severity (e.g. SEV1: status-page + update every 30 min; normal: reply within SLA). Who owns the customer comms vs. the fix.
test cases passed, vs 18 / 22 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.