Knowledge OPS
Knowledge OPS is a skill for authoring, validating, and cleaning up company SOPs and internal runbooks. It provides deterministic Python tools to check 5W2H completeness, validate runbook steps against six required attributes, and audit knowledge bases for orphan pages, stale content, glossary drift, and missing owners, producing prioritized cleanup lists.
✨ What it does
- Generates 5W2H-structured SOP scaffolds with profiles for ops, support, finance, hr, it, and regulated (with compliance
- Validates runbook steps against six checks: named owner, expected duration, observable success/failure signals, rollback
- Audits KB exports for orphan pages, stale pages (no edit > 12 months), glossary drift, missing-owner pages, and cross-li
- Produces a prioritized top-20 cleanup list ranked by staleness × inbound-link-count.
- Includes sample modes for all scripts to demonstrate functionality.
- Provides reference documents on SOP authoring canon, runbook design, and KB hygiene anti-patterns.
🎯 When to use it
- Authoring a new SOP for a cross-functional process like procurement intake, vendor offboarding, or incident-comms cascade.
- Validating an existing runbook before it goes into rotation to ensure every step has an owner, duration, success/failure signals, rollback p
- Ingesting a multi-document KB export (Notion, Confluence, Obsidian) to surface broken items like orphan pages, stale pages, and glossary dri
- Onboarding a new ops hire by generating the SOPs and ops-handbook pages they need to read in week 1.
- Running a wiki cleanup sprint to decide which docs to archive, rewrite, or merge.
🚀 How to use
Trigger: /healthz
Trigger by asking to author, validate, or clean up SOPs/runbooks, e.g., 'Validate this runbook before it goes into rotation' or 'Audit our Confluence wiki for stale and orphaned SOPs'. Provide a markdown file or directory path for validation/ingestion, or a JSON metadata file for SOP generation. Use the scripts with appropriate flags: python3 scripts/runbook_validator.py --input <runbook.md>, python3 scripts/kb_ingester.py --input <vault-dir>, python3 scripts/sop_generator.py --input <metadata.json> --profile <profile>. Example prompts:
Validate this runbook before it goes into rotation: /path/to/runbook.md
Audit our Confluence wiki export for stale and orphaned SOPs: /path/to/export/
📄 Output: The skill produces markdown reports (KB health report, runbook validation report) and SOP scaffolds in markdown or JSON.
📦 Add this skill to Claude Code
# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills
# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/business-operations/skills/knowledge-ops .claude/skills/knowledge-opsSkill source: business-operations/skills/knowledge-ops/SKILL.md
⚠️ Good to know
Requires the KB to be in markdown (or exportable to markdown); HTML/PDF-only KBs need a conversion pass. The user must have authority to act on cleanup lists, and owner metadata must be in YAML frontmatter or a top-of-page 'Owner:' line.
❓ FAQ
What does the runbook validator check?
It checks each step for six attributes: named owner, expected duration, observable success signal, observable failure signal, rollback path, and escalation contact. It outputs a traffic-light per step and an overall validity score.
How does the KB ingester detect stale pages?
It uses filesystem mtime or YAML last_reviewed frontmatter to find pages with no edit in over 12 months (configurable with --stale-days).
What is the difference between knowledge-ops and process-mapper?
Process-mapper documents the flow of work between stages (BPMN, cycle time), while knowledge-ops documents the artifacts operators consume to execute the work, like SOPs and runbooks.
🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.