Your library
SkillsYour company's knowledge, loaded where neededPersonasRoles that know what to loadPromptsThe tasks you repeat, ready to fill inGovernanceWho sees, who edits, review, versionsUse it in
Browser extensionPlaybooks inside ChatGPT, Claude, Geminisqemes ChatAny provider you connect, with personasMCP serverClaude Code, Cursor, Codex and more
Probably AI-generatedSee it in Claude CodeHow an agent finds and loads a skill
By team
MarketingOne brand voice in every toolSalesYour best messaging in every replyEngineeringStandards your agents can followProductProduct context you can build onSupportOne source of truth for every answerOperationsRecurring work, actually repeatableBy role
Team leadsShared playbooks, clear rightsEnterpriseSelf-host or a dedicated instancePartnersFor agencies and implementers
Probably AI-generatedOne library, every teamHow teams share playbooks
Builds a rich text-to-image prompt plus a negative prompt.
Fields you fill in
{{subject}}Subject
{{style}}Style
{{mood}}Mood
Build a detailed text-to-image prompt.
Subject: {{subject}}
Style: {{style}}
Mood: {{mood}}
Return one rich prompt that specifies composition, lighting, color palette, camera or medium, and level of detail. Then add a short negative prompt of things to avoid.