Critiquito, by Manuel Muñoz Solera, turns a product screen into ranked design findings with concrete fixes. The official Marketplace calls it "Critiquito: Design Critique" and positions it around practical usability feedback rather than creating a new visual identity.
Last checked: November 6, 2026 How we review: based on the official listing and its published instructions; we don't install bots or connect accounts.
Open the official Critiquito listing.
What it does
You supply a screenshot, a Figma frame, or a live URL. Critiquito returns a verdict and findings across hierarchy, spacing, type, color, interaction and states, copy, and accessibility. Each finding should have a specific fix and one of three severities: blocker, worth fixing, or minor.
The instructions split work by input. One screen goes to a screen critique; two versions go to a comparison; a sequence goes to a flow critique. Contrast and target questions go to an accessibility pass, while labels and errors go to a UI-copy pass. There is also a handoff path for tickets or an archive page.
Its working files hold a dated critique for each screen and an open fix list. When a later version arrives, the bot is instructed to reread that list and check whether the old findings cleared instead of repeating them blindly. Ticket or archive links are recorded with the screen's findings.
Who it is for
Product designers, founders, and product managers can use it to prioritize what to fix in an existing screen. It is especially relevant when a team has variants to compare or a short flow whose friction needs a concrete explanation.
It is not meant to design screens for you, create illustrations, or do brand work. Bring the screen and explain who uses it; do not expect a screenshot alone to communicate business rules, hidden states, or user research.
Integrations and inputs
Screenshots, Figma frames, and live URLs are named inputs. The public listing does not provide a complete connector list, Figma scopes, ticketing-system names, or channel integrations. A handoff instruction does not establish that your preferred issue tracker is connected.
Getting-started preferences include the product, its users, platform, severity bar, design-system or brand rules, where fixes should go, timezone, and the weekly check day and hour. Existing preferences let it skip those questions and ask for the screen.
The page says it should put a critique in front of you within a minute. That is an instruction about response behavior, not an installation-time estimate; no setup duration is given.
Routines
One routine is named: the weekly open fixes check. It stays off until enabled and runs in your timezone. Set the day and hour deliberately, and confirm which screen lists it reads and where it reports findings before turning it on.
Keep the first critique manual. A stale fix list can produce stale advice, so supply the latest version and make clear which screen or flow replaces the earlier one.
Safety check
Safety: Critiquito reads the screenshots, Figma frames, or URLs you share and keeps critique files; its instructions require your yes before design edits, ticket creation, channel posts, or page publishing, so check file access, destinations, permissions, and the disabled weekly check in the Add preview.
Its core job is critique, not editing. The broader approval instruction is a reason to inspect external-action access, not evidence that every named destination has a connector. Keep the initial result in the conversation and approve a specific handoff separately.
Use screens you are authorized to share. Customer details, internal URLs, and unreleased product plans can all appear in screenshots. Remove anything the critique does not need. The bot's stated rules describe intended behavior; check the actual permissions rather than relying on the wording alone.
Step-by-step installation
- Open the official page and confirm the creator is Manuel Muñoz Solera.
- Choose Import Bot, Add on the Marketplace card, or the Add to Grok Bot action described in the official guide.
- Review context and integrations in the preview. Look specifically at design-file access and any ticket, page, or messaging permissions.
- Choose Add Bot after reviewing that access. This creates a separate copy, not a clone of the creator's environment.
- Reinstall any required plugins. Set the product, audience, platform, severity bar, and design-system rules; leave the weekly check off initially.
- Share one non-sensitive screen and ask for findings in chat. Review the fixes before authorizing any external handoff.
Three example tasks
- Review a screen: "Critique this signup screen for first-time users on mobile. Lead with the verdict, rank findings by severity, and give a concrete fix for each. Do not edit or file anything."
- Compare versions: "Compare these two checkout variants for the same task. Explain which visible changes reduce friction and what remains uncertain from the screenshots."
- Check a revision: "Review this updated screen against its open fix list. Mark which findings cleared and which remain, then draft a handoff here without creating tickets."
Limitations and when not to use it
The instructions say to judge only what is visible, avoid inventing measurements or hex values, and flag images too compressed to measure. Preserve those limits: a screenshot is not a full accessibility audit, a keyboard-navigation test, or proof that the underlying implementation behaves correctly.
Do not use it as a substitute for user research or brand direction. For exact implementation values, provide the design source and consider a spec-focused workflow. For changes in a real file, use explicit approvals and review the result yourself.
Alternative templates
- figma bro fits when you need build specs, component and token audits, or a screen built from a brief rather than critique alone.
- Site Audit fits a broader website review across search, content, speed, accessibility, conversion, and schema rather than one product screen.
FAQ
Does Critiquito edit my Figma files?
The listing positions it as a critique bot, not a design editor. Its instructions require your yes before editing a design file, filing a ticket, posting to a channel, or publishing a page.
What does it review?
It reviews screens, variants, and short flows across hierarchy, spacing, type, color, interaction states, copy, and accessibility. Findings use blocker, worth fixing, or minor severity.
Is the weekly check enabled by default?
No. The weekly open fixes check stays off until you turn it on and runs in your timezone. Its exact schedule is a setup preference.
Compare the video & creative templates, or return to the Grok Bot templates hub.