Change & Problem Management

15 skills in this category.

CAB Brief Builder

Build the weekly change-board pack — pending changes ranked by risk, collision flags, and last week's change outcomes — so the CAB meeting is 20 minutes of decisions instead of an hour of reading tickets aloud.

Change Calendar Management

Check a proposed change window against freeze windows, client business calendars, and other scheduled changes on the same systems — catch the collision before the maintenance email goes out, not during the outage.

Change Request Intake

A change ask arrived as prose ("can we upgrade the firewall this weekend?") — normalize it into a structured change request (what/why/scope/when/rollback/risk) and route it to the right approval track before anyone touches anything.

Change Risk Assessment

Classify a change request as standard, normal, or emergency by scoring blast radius and rollback confidence — so approval effort matches actual risk instead of everything getting the same rubber stamp.

Emergency Change Handling

A change had to happen NOW to restore or protect service — run the break-glass discipline: minimal in-flight record, act-then-document is allowed, but the full retro-documentation is mandatory within 24 hours and this skill chases it to done.

Incident Commander Brief

An incident commander is taking over a running major incident — assemble everything they need in one read: timeline so far, workstream states, comms state, and the next decision point. Takeover without the 20-minute "so what happened?" retelling.

Incident Comms Cadence

During a major incident, updates go out on a clock — internal and client — even when the update is "no change." Draft each update from ticket evidence, track the next-due time, and never let the cadence silently lapse.

Known Error Database

Maintain the KEDB — every known error in one findable format (symptom / root cause / workaround), deduplicated on arrival, and retired when the fix ships — so techs stop re-diagnosing solved mysteries at 2am.

Maintenance Freeze Windows

Record and enforce client freeze calendars (tax season, go-lives, retail peak) — freezes block change scheduling at intake, and the only way through one is the documented exception path with the client's named sign-off.

Major Incident Declaration

Something big is breaking — run the criteria check, declare (or explicitly don't), assign the incident roles, and start the comms clock. The declaration checklist that turns "everyone panicking in chat" into a managed major incident.

Post-Incident Action Tracking

Turn post-incident review action items into real tickets with owners and due dates, then run the follow-through audit that catches the ones quietly dying — the discipline that makes postmortems more than a feelings meeting.

Problem Record Lifecycle

Drive a problem record through its states — opened from an incident cluster, under investigation, known error, and finally fixed or accepted-risk closure — so problems either get solved or get a deliberate decision, never a quiet death in the backlog.

Recurring Maintenance Tickets

Scheduled maintenance tickets (backup checks, patch cycles, monthly server reviews) rot into checkbox theater — verify each cycle carries real completion evidence and flag skipped cycles before "monthly maintenance" quietly becomes quarterly.

RFO Letter

Draft the reason-for-outage letter a client receives after a major incident — facts, impact, remediation, prevention — written defensively: every sentence is one the MSP can stand behind if the letter ends up in front of lawyers or insurers.

Workaround Documentation

A workaround living in one tech's head is a liability — document it in the standard format (steps, hold time, cost, expiry review), label the ticket "workaround-only," and make sure nobody mistakes it for a fix.