Where product teams and agents see how competitors build.
The research library of real, shipped fintech apps in Uzbekistan and Georgia. Every screen, flow and form rule, taken from the actual release build, with the evidence attached.
Market research took weeks. Now it's one search.
Every PM has that deck: fifty screenshots of competitor apps, taken on a borrowed phone, out of date by the time the review happens. Specimen replaces it with a library that is complete, current and searchable.
Not screenshots. Answers with receipts.
Ask how a competitor handles a limit, an error, a confirmation. The answer cites the screen, the string resource and the code path it came from, so it survives a design review.
Payments tab → category → a server-driven service form (the bundled form JSON: labels, masks, min/max) → a confirmation screen with the details, the debit card and a bottom-sheet card picker → a payment status dialog. Providers change without an app release.
// claude_desktop_config.json { "mcpServers": { "specimen": { "url": "https://mcp.specimen.app/v1", "headers": { "Authorization": "Bearer sk_live_…" } } } } // then, in your agent: // "Compare how Click, Payme and TBC UZ confirm // a P2P transfer. Table with evidence."
Give your AI agents access to competitors.
Your PRD agent, your design agent, your Cursor session: all of them can search screens, pull flows and compare features over MCP. Available on Team and Enterprise.
Plans
Per seat, billed monthly. Apps are shared across your workspace.
- 3 apps from the library
- Screens, flows, design tokens
- Ask with evidence
- Export to Markdown
- Unlimited apps in 2 markets
- MCP · 5 000 calls / mo
- Clickable prototypes
- Release diffs on every update
- Shared collections
- All markets, custom app requests
- Unlimited MCP, private endpoint
- Release-day teardown SLA
- SSO, audit log, data residency