sysadmin-skills

Use this skill whenever the user wants change-management paperwork (RFC, change request, change record, governance doc, rollback doc, CAB submission) for anything AI or ML. This is the correct skill — not a generic RFC skill — any time the subject of the change is a model, LLM, chatbot, AI agent, prompt, RAG pipeline, embedding, knowledge base, training data, or fine-tuning job. Common situations: swapping LLM providers or model versions, deploying or retraining an ML model, RLHF or fine-tuning runs, editing production prompts, refreshing RAG sources, granting an agent new autonomous actions or tools, and emergency rollbacks when an AI is hallucinating or producing harmful output. Rule of thumb: if the user is asking for approval/governance paperwork AND the thing being changed could shift an AI's behavior, this skill wins over any general change-request skill. Also use for questions about ITIL v5 AI Governance or the 6C model (Creation, Curation, Clarification, Cognition, Communication, Coordination).

ITIL change request and CAB review workflow. This skill should be used when the user asks to "create a change request", "write an RFC", "plan a change", "CAB review", "change management", "schedule maintenance", "server migration plan", "upgrade plan", "AI model deployment", "GitOps change", or describes any planned modification to IT infrastructure or services. Produces a complete RFC document following ITIL v5 Change Enablement.

Operations task verification gate before claiming work is complete. This skill should be used before closing incidents, finalizing change requests, sending resolution notifications, or when asking "is this resolved?", "can we close this?", "is the change complete?", or verifying SLO compliance after an incident. Evidence before claims — verify operational status first.

IT incident response workflow. This skill should be used when the user asks to "log an incident", "handle an incident", "create incident report", "document an outage", "incident response", "service is down", "major incident", or describes a service disruption requiring structured response. Also use when the user mentions SRE concepts like error budget impact, blameless review, or AIOps-detected incidents. Guides through the full ITIL v5 incident lifecycle: detect, classify, investigate, resolve, communicate, and close.

ITIL (Version 5) framework reference for IT service management and digital product management. This skill should be used when handling incidents, creating change requests, conducting CAB reviews, classifying priority levels, asking about ITIL v5 practices, AI governance, SRE integration, error budgets, or sustainability in IT operations. Provides Incident Management, Change Enablement, Priority Matrix, AI Governance (6C model), and SRE guidance.

Use this skill when the user wants help writing up a resolved production incident, outage, or failure. This is the go-to skill for any request that involves turning "something broke and we fixed it" into a written document — whether the user calls it a postmortem, post-incident review, post-incident analysis, PIR, incident retrospective, blameless review, no-blame writeup, lessons learned, or just describes what broke (deploy, database, auth, payment, storage, backup, replication, service outage) along with details like severity, duration, users affected, or rollback, and asks for help documenting it. Also use when the user's goal is team learning or capturing details before they're forgotten from a past incident. Output is a blameless document with timeline, contributing factors, what went well, and action items. Do not use for active/ongoing incidents, sprint or agile retrospectives, code reviews, general project retros, problem records for recurring issues, change requests, or meeting agendas.

Use this skill when the user wants to open a problem record, register a known error (KEDB), or run root cause analysis (RCA, 5 Whys, Fishbone) on an IT issue. Trigger on explicit phrases: "problem record", "problem ticket", "known error", "KEDB", "RCA", "root cause analysis", "systemic weakness", "latent risk". Also trigger when the user describes a recurring failure ("every Monday", "every Saturday night", "keeps happening weekly", "5 incidents this month all pointing at..."), wants to proactively track a trending risk before it becomes an outage ("disk trending to 100%", "capacity heading bad", "no smoke tests yet"), or needs to link a workaround to a permanent fix. The user's intent is to track and eliminate an underlying cause across multiple occurrences — not to fight a live outage, not to write a postmortem for a single past incident, and not to draft a change request. If they say "problem", "known error", "recurring", or "systemic" in an ITSM context, use this skill.

Stakeholder communication templates for IT operations. This skill should be used when writing incident notifications, status page updates, maintenance announcements, or any communication to users, management, or other teams about IT service events. Provides severity-based tone guidance and audience-aware messaging.

Status page update workflow for IT service incidents and scheduled maintenance. This skill should be used when the user asks to "write a status update", "update the status page", "post maintenance notice", "draft statuspage update", "service status", "notify users about outage", or needs to communicate service status to users. Produces updates following the Investigating/Identified/Monitoring/Resolved lifecycle, compatible with Atlassian Statuspage and Instatus.