01 Browse
Search by intent, not filename.
Queries match outcomes, variables, tags, and examples. Category shelves stay visible so users can pivot without losing context.
Prompt library for product teams
Find the prompt worth running.
Shelfmark turns scattered team prompts into a browsable catalog with tested examples, model fit notes, and reusable prompt paths.
486curated prompts
92%reviewed monthly
14ksaved runs
No prompt shelf matches that path.
Try clearing one filter or searching for research, memo, interview, or roadmap.
Try clearing one filter or searching for research, memo, interview, or roadmap.
Credible prompts, not a folder of guesses.
Every entry carries enough operating detail for a senior designer, PM, or researcher to judge whether it belongs in a real workflow.
Prompt
Reuse
Tokens
Fit
Interview Guide for Risky Assumptions
1.8k
Deep
Competitor Claim Extractor
940
Fast
Roadmap Tradeoff Brief
2.4k
Deep
Shelfmark made our prompt library feel like a product surface. The team can browse by job, inspect evidence, and chain a workflow without asking where the good version lives.
Maintainers see stale prompts first: low reuse, older model fit notes, missing examples, or failed copy events.
A browse rhythm built around real use.
The catalog starts with the job-to-be-done, then reveals prompt evidence only when it helps a user decide.
02 Inspect
Open the operating notes.
Each prompt includes model fit, expected tokens, example output, failure cases, and the last reviewer.
03 Chain
Save a reusable path.
Selected prompts become an Intent Path with variable slots, order, and handoff instructions for teams.
Start with a maintained library.
For design and product teams turning ad hoc AI usage into a shared operating layer.
$18 per editor / month
- Unlimited read-only teammates
- Prompt review queues
- Intent Path builder
- Model fit notes
- Usage and copy metrics
- CSV import from docs
Shelfmark