Agent
Skills
One public, self-configuring Agent Skill per level. Each asks what you do, then adapts — so the same skill works for any role. Install one, or grab the whole set.
prompt-architect/
Turn a vague request into a reliable, reusable 5-part prompt. It asks your role and task, then builds it — for any field.
View on GitHubcontext-brain/
Build a reusable context pack so Claude reasons from your world — role, audience, voice, standards — instead of generically.
View on GitHubskill-author/
Package a task you repeat into your own SKILL.md skill, so it runs the same every time instead of depending on memory.
View on GitHubconnector-setup/
Plan a safe, least-privilege way to connect your real tools to Claude so it works from live data, not copy-paste.
View on GitHubpipeline-builder/
Break a big, multi-part task into a clean pipeline where each step's output feeds the next, with context carried forward.
View on GitHubwork-engine/
Stand up a repeatable brief → deliverables → report → learnings loop for any work you do on a cadence.
View on GitHuborchestrator/
Assemble your context brain, skills, connectors, and pipelines into one system you actually run — and document it.
View on GitHub