Customer Service & Retention AI

Knowledge Ops

Knowledge Ops is an ecommerce AI skill for Alireza Rezvani, built for teams working with Codex, Claude Code, OpenClaw. Use it to you are deciding what to improve from…

Provider
Alireza Rezvani
Platforms
Codex · Claude Code · OpenClaw
View original link · GitHub

What this skill helps you do

Addresses the sprawl of company wikis, SOPs, and runbooks that accumulate over years. Validates runbooks against six required attributes per step (named owner, expected duration, success signal, failure signal, rollback path, escalation contact) and scores them 0-100 with a NOT-SAFE verdict below 60. Ingests multi-document KB exports to find orphan pages, stale docs, glossary drift, and missing-owner pages, producing a prioritized top-20 cleanup list. Generates 5W2H-structured SOPs with regulatory overlays (SOC2, HIPAA, ISO 13485, GDPR, SOX) and six profiles from general ops to regulated.

Install and get started

Copy the full instructions into your AI tool. Test one low-risk example before connecting real store data.

Original Skill instructions

You are a Knowledge Operations specialist. Three deterministic Python tools: (1) kb_ingester.py on KB export—ranks top-20 docs to fix, flags orphan pages, stale pages (>12mo), glossary drift, missing owners, cross-link gaps. (2) runbook_validator.py on each runbook—scores steps against 6 checks: named owner, expected duration, observable success signal, observable failure signal, rollback path, escalation contact. ≥80=SAFE, <60=NOT-SAFE. (3) sop_generator.py with --profile (ops|support|finance|hr|it|regulated)—produces 5W2H SOPs with regulatory overlays. Never bulk-generate SOPs without named owners. Fix glossary drift immediately. Metrics that matter: unfindable docs and unsafe runbooks.

Useful tasks

  • Quarterly wiki cleanup sprints with prioritized top-20 doc fix list
  • Validating incident runbooks before they go into rotation
  • Generating compliance-ready SOPs with SOC2/HIPAA/GDPR overlays
  • Onboarding new ops hires with curated reading lists from KB inventory
  • Detecting glossary drift where the same acronym means different things across teams

How to use it

  • Run kb_ingester.py before every cleanup sprint—the top-20 list tells you what to fix first
  • A runbook scoring <60 is NOT-SAFE: don't let it into an incident rotation
  • Use --profile regulated for any SOP touching PHI, SOX controls, or ISO 13485 QMS
  • Fix glossary drift immediately—two definitions today becomes seven in five years
  • Route findings to a named owner before running the ingester

More skills for this workflow

Content checked: