Where your team's AI work lives.
A registry for the prompts, skills, agents, and workflows your team writes. Connect once — your agent reads what it needs from Claude, ChatGPT, Cursor, or Codex. No installs, no copy-paste.
Free for crews of 5. No card.
Top resources on the bench
Skill Creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Bash Pro
Master of defensive Bash scripting for production automation, CI/CD pipelines, and system utilities. Expert in safe, portable, and testable shell scripts.
Documentation
Creates, structures, and reviews technical documentation following the Diátaxis framework (tutorials, how-to guides, reference, and explanation pages). Use when a user needs to write or reorganize docs, structure a tutorial vs. a how-to guide, build reference docs or API documentation, create explanation pages, choose between Diátaxis documentation types, or improve existing documentation structure. Trigger terms include: documentation structure, Diátaxis, tutorials vs how-to guides, organize docs, user guide, reference docs, technical writing.
Bats Testing Patterns
Master Bash Automated Testing System (Bats) for comprehensive shell script testing. Use when writing tests for shell scripts, CI/CD pipelines, or requiring test-driven development of shell utilities.
Save it once. Read it from anywhere.
Connect your bench once on the agent side. The tools your team writes in publish straight to the bench — and any tool can read them back, same version, same author, on demand.
How it works
Share once.
One person on your team writes a brand-voice agent, a sales-email skill, or a customer-research workflow.
Use anywhere.
The rest of the team pulls it from any AI tool — Claude, ChatGPT, Cursor, Codex — one click later. Every pull is the version the author shipped today.
Team picks it up.
Your work outlives the worker. New hires onboard against your team's actual AI craft — and they know who wrote it.
Common questions
What is Sharebench?
Sharebench is a registry for AI prompts, skills, agents, and workflows. Your team publishes once; everyone uses them from Claude, ChatGPT, Cursor, or Codex over MCP.
How do I make sure my team uses the latest version?
Every pull reads the version the author shipped today. When the author updates the prompt, the next agent run picks up the update — no notification to action, no stale copy in someone's Notion.
Can I share the same prompt across Claude and ChatGPT?
Yes. Save it once on Sharebench. Your agent reads it from Claude Code, Cursor, ChatGPT, or Codex — the format adapts per tool.
Where does the prompt live? Do I install anything?
Nothing to install. Your agent pulls the prompt on demand over MCP. Connect once on the agent side and your team's bench is available.
Who gets credit for a prompt I publish?
You do. Your name stays on the artifact through every pull, save, and reuse — including when teammates onboard against your work.