Guides
How to keep AI skills consistent across a team and across tools
The four ways teams share AI skills across Claude, ChatGPT, Cursor, and other tools, and how each handles updates, editing, and drift.
SkillRelay vs. Claude org-provisioned skills
When Anthropic's built-in org skill provisioning is enough, and when a cross-tool skill library is worth running alongside or instead of it.
SkillRelay vs. a shared Git repo of skills
A Git repo is a solid skill store for engineering teams. Where it breaks is edit access: anyone not in a terminal can read skills but never change them.
What is an AI skill (and a SKILL.md), for non-coders
A skill is a short instruction document an AI tool reads before doing a task. SKILL.md is the standard file format skills are stored in.
Versioning, authorship, and approval gates for AI skills
How SkillRelay records every skill change as an immutable, attributed version, and how optional approval gates work.
How to share AI skills across a team without engineers
A team with no engineering support can run a shared AI skill library: what the non-technical path looks like and what to avoid.
From a shared prompt doc to a team skill library
Why teams outgrow the shared prompt document in Notion or Google Docs, and what replaces it: skills that load themselves.
Who owns a team's skill library
Ownership of a shared skill library: roles, offboarding, data custody, and how to leave with everything.