--- title: "Judge.me Integration" description: "Work product reviews in Judge.me from your flows. Agents list, create, update, and moderate Shopify store reviews, look up products, read total and per-product review counts with average ratings, and register webhooks for review events." url: https://flowrunner.ai/integrations/judge-me date_modified: 2026-08-04T10:04:16-07:00 --- # Judge.me [E-commerce](https://flowrunner.ai/integrations/category/e-commerce) Work product reviews in Judge.me from your flows. Agents list, create, update, and moderate Shopify store reviews, look up products, read total and per-product review counts with average ratings, and register webhooks for review events. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 11 actions · API key · available [Judge.me website](https://judge.me/) · [Platform Documentation](https://judge.me/api/docs) · Capability data verified 2026-07-27 1. The nightly review sweep starts after the store's day closes 2. List Reviews pulls the day's new reviews with ratings, reviewer details, and product associations 3. Get Product resolves each review's product handle and title so the report reads like the catalog, not like IDs 4. The agent flags one- and two-star reviews, and separates product complaints from shipping complaints by what the text actually says 5. Flagged reviews post to the support channel with the product, rating, and full review body 6. The store owner decides each moderation call: reply publicly, contact the customer, or change the published state 7. Update Review applies only the approved published-state changes, nothing on the agent's own judgment 8. Get Product Review Count refreshes the average rating and count feeding the marketing dashboard ## What This Integration Enables Reviews are the highest-converting content on a storefront and usually the least governed: collected by one app, read occasionally, moderated by mood. Judge.me gives Shopify stores the collection machinery; FlowRunner gives the operation around it. Agents list and retrieve reviews with full reviewer and product context, import reviews gathered on other channels, update ratings, titles, bodies, and published state, resolve products by handle or ID, and read the aggregate counts and averages that power social proof. - Triage: nightly sweeps that sort new reviews by severity and route the ones needing a same-day response - Sync: review streams into spreadsheets, warehouses, and CRM timelines where the rest of the company can see them - Import: reviews collected off-platform brought into the product's history, deliberately and once, because Judge.me does not deduplicate - Social proof: live counts and average ratings feeding dashboards, email campaigns, and on-site claims The line that matters: what shoppers see is a publishing decision. Agents read, sort, and draft at machine speed; the changes that alter the storefront's public record run through [human-in-the-loop](https://flowrunner.ai/concepts/human-in-the-loop) approval. ### Without FlowRunner **Reviews get read on Fridays**: New reviews pile up unread, and the one-star that needed a same-day reply waited four **Social proof is a screenshot**: Review counts and averages get copied into decks and emails by hand, already stale **Moderation is invisible**: What got hidden, when, and why lives in no record anyone can check ### With FlowRunner **Every review is triaged by morning**: Agents sweep nightly, sort product issues from shipping issues, and wake the right person for the bad ones **The metrics feed themselves**: Counts and averages flow to dashboards and campaigns straight from the widget data **Moderation leaves a trail**: Publish-state changes happen only on recorded human approval, with the reasoning attached ## Use Case Scenarios ### The one-star that gets answered before breakfast The nightly sweep runs List Reviews and finds a one-star review on the store's best seller. The agent pulls the reviewer's recent order from [Shopify](https://flowrunner.ai/integrations/shopify), sees a delivery delay, and posts the whole picture to [Slack](https://flowrunner.ai/integrations/slack): review text, rating, order timeline, and a drafted response acknowledging the late shipment. The owner edits two words and replies within the hour. The review stays published, the reply shows future shoppers a store that answers, and nobody hid anything. ### Reviews from everywhere, imported once The brand collects feedback through a post-purchase survey in [Typeform](https://flowrunner.ai/integrations/typeform) and gets product praise in support tickets. Qualifying submissions with explicit permission queue for import, matched to products through Get Product By Handle. Because Judge.me does not deduplicate imported reviews, the agent checks List Reviews for the same reviewer and product before anything is written, and a person approves the batch. Create Review then imports each one with the reviewer's name, rating, and text. Off-platform praise becomes on-platform proof without double-posting a single entry. ### Social proof numbers that are never stale Every morning the agent calls Get Total Review Count for the store and Get Product Review Count for the hero products, writing counts and average ratings to [Google Sheets](https://flowrunner.ai/integrations/google-sheets) and pushing profile events to [Klaviyo](https://flowrunner.ai/integrations/klaviyo) for products crossing rating thresholds. The email campaign claiming "4.8 stars across 2,000 reviews" reads those numbers from the feed, not from a screenshot taken in March. When a product's average slips below the threshold the team set, the drop itself posts to the channel with the recent reviews that caused it. ## Human-in-Loop Highlight Update Review can flip a review's published state, and that flip edits the storefront's public record. Unpublishing a negative review is sometimes right, spam, harassment, a competitor's plant, and sometimes it is quietly deleting the most honest thing a customer ever told you, with platform review policies watching either way. That judgment does not belong to an agent. In FlowRunner workflows, the agent flags the review, gathers the order context, and drafts the case for both directions; a person makes the call, and Update Review executes exactly that decision with the reasoning logged. Create Review carries the same weight in the other direction: Judge.me does not deduplicate imports, so an unattended import loop can stamp the same review onto a product three times, which reads as fake because it is. Batches import once, checked and approved. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via preferred channel Human decides Agent resumes with decision ## Agent Capabilities 11 actions ### Reviews 4 - **List Reviews** Lists the store's reviews newest first, filterable by product, star rating, published state, and reviewer, up to 100 per page. The read behind every sweep and sync. - **Get Review** Retrieves one review by ID with rating, title, body, reviewer details, product association, and published state. - **Create Review** Imports a review for a product with reviewer name, email, rating, and body, matched by Shopify product ID. No deduplication on Judge.me's side, so match first and import once. - **Update Review** Updates a review's rating, title, body, or published state; only supplied fields change. The published-state flip is the reason this page has a human gate. ### Products 3 - **List Products** Lists the store's products known to Judge.me with internal ID, Shopify external ID, handle, and title. - **Get Product By Handle** Looks up a product by its Shopify handle, the URL slug, returning its IDs and title. The matcher for imports keyed on storefront URLs. - **Get Product** Retrieves one product by Judge.me internal ID with its external ID, handle, and title. ### Widgets 2 - **Get Total Review Count** Returns the store's total published review count, the headline social-proof number. - **Get Product Review Count** Returns a product's review count and average star rating from its widget, identified by handle. The per-product number campaigns quote. ### Webhooks 2 - **List Webhooks** Lists the webhooks registered for the store with topic and callback URL. - **Create Webhook** Registers a callback URL for a review event, such as review created, so downstream workflows start the moment a shopper hits submit. ## Frequently Asked Questions ### What can FlowRunner do with Judge.me? FlowRunner agents can run List Reviews, Get Review, and Create Review in Judge.me, plus 8 more actions. ### Does connecting Judge.me to FlowRunner require OAuth? No. Judge.me connects to FlowRunner with an API key, no OAuth flow required. ### Can Judge.me trigger a FlowRunner workflow automatically? Judge.me doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at Judge.me?** This integration exposes Judge.me to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current. --- Markdown version of https://flowrunner.ai/integrations/judge-me. Site index: https://flowrunner.ai/llms.txt