Team Prompt Library Governance
Organize shared prompts so they stay good.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
[VERIFY] is an editorial flag, not a blank to fill. Wherever the prompt shows [VERIFY], check that item against a reliable source before you use it.
Design governance for a team prompt library. Team, size, and uses: [TEAM_AND_USES]. Tools where prompts live: [TOOLS]. Current problems (duplicates, outdated prompts, untested prompts, sensitive data in prompts): [PROBLEMS]. Risk level of the work: [RISK_LEVEL]. Provide: 1. Library structure: categories, naming, tags, owners, and a template for each entry (purpose, inputs, prompt text, example output, limits, risks, tested date, version). 2. Roles: contributors, reviewers, maintainers; who can publish; how to propose and retire prompts. 3. Quality gate: before publication, a prompt needs a purpose, at least 3 test cases, a review by another person, and a note on risks; high-risk prompts (customer-facing, legal, HR, financial) need extra approval. 4. Data rules: no personal data, secrets, or confidential client information inside prompts or examples; approved tools only. 5. Review cycle: reviewing top-used prompts every quarter and removing unused prompts. 6. Feedback loop: ratings, comments, issue reporting, and usage statistics (without monitoring individuals unnecessarily). 7. Onboarding: a one-page guide for new users and a training session outline. 8. Metrics: usage, satisfaction, time saved (self-reported), incidents. Include a template entry filled for a meeting summary prompt, and a policy statement on responsible use. Have security and legal teams review the data rules [VERIFY].
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.