Papercuts

About Papercuts

Papercuts is a testing and QA tool that runs AI agents through your live application so they interact with the interface like a human user. Point the service at a URL and the agents visually explore flows, reporting when user-facing UI or logic breaks occur.

Review

Papercuts fills a gap between script-based tests and traditional monitoring by using vision-based agents to detect regressions that DOM checks can miss. It is intended to run in production alongside existing tools, providing early alerts for subtle UI and flow issues before they reach users.

Key Features

• Vision-driven agents that interact with the page visually rather than relying on fragile DOM selectors.
• Continuous exercise of production flows to detect UX and logic regressions in real contexts.
• Notifications when interactions fail or when visual/behavioral changes break a flow.
• Adaptive handling of dynamic or conditional form fields by observing visible elements and responding accordingly.
• Integrates conceptually with existing test suites and monitoring, acting as an additional layer of user-focused checks.

Pricing and Value

The product offers a free plan to get started and encourages trying agents on a few critical flows before widening coverage. For teams that repeatedly see "green" tests but still get user complaints, Papercuts provides value by surfacing hard-to-script problems earlier, which can reduce support load and user friction. Exact paid tiers and usage limits are not detailed here and should be reviewed on the provider's site for production-scale needs.

Pros

• Finds UI and conditional-flow issues that DOM-based tests can miss by simulating human interaction.
• Runs in production so checks reflect real application state and external integrations.
• Handles dynamic forms and conditional fields by observing visible context.
• Can complement existing Playwright/Cypress suites and APM/RUM rather than replace them.
• Quick to start: provide a URL and agents begin exercising flows.

Cons

• Running AI agents in production may require careful configuration to avoid noisy alerts or unintended side effects.
• Some tuning is likely needed to reduce false positives and tailor checks to a team's critical flows.
• Not a substitute for deterministic, scripted tests where exact assertions and reproducibility are required.

Overall, Papercuts is best suited for product and QA teams that operate complex user flows-especially multi-step forms or conditional interfaces-and who already use scripted tests and monitoring. It works well as an additional safety net to catch user-visible regressions that other layers miss, while teams validate whether agents should cover more production paths over time.

Open 'Papercuts' Website