---
title: "Curator — Spots overlapping articles and recommends fixes"
description: "Reviews your whole blog for cannibalization and low-value pages, then recommends consolidations for you to approve — it never changes anything on its own."
url: "https://skribra.com/agents/curator"
type: "agent"
kind: "recommends (waits for approval)"
---

# Curator

**Spots overlapping articles and recommends fixes.** Reviews your whole blog for cannibalization and low-value pages, then recommends consolidations for you to approve — it never changes anything on its own.

## At a glance

- **Signal it watches** — Clusters of published articles competing for the same search (cannibalization), plus thin, dated, or duplicate pages — with Search Console used to pick which article in a cluster is the keeper.
- **What it does** — Writes recommendations — recreate with a better title, 301-redirect a duplicate into the keeper, noindex, or remove — and waits for your approval. Nothing changes until you say yes.
- **Cadence** — Reviews each project about once a month (admins can run it on demand). Approved rewrites draw a small monthly quota; redirects and removals don't.
- **Autonomy** — proposes and waits for your approval

## What it is

Publishing at volume eventually creates a quieter problem: your own articles start competing with each other for the same search, splitting clicks and confusing Google about which page to rank. Curator is the agent that cleans this up — but unlike the others, it only recommends.

Every month it groups your published articles by real topic and intent (collapsing near-synonyms like "rank tracker" and "google website ranker"), picks the strongest page in each cluster by Search Console performance, and proposes what to do with the rest: rewrite a mis-titled page, [301-redirect a duplicate](https://skribra.com/glossary/301-redirect) into the keeper so its ranking signals transfer, or retire dead weight.

You review each suggestion in the dashboard and approve the ones you want. Recreates keep the original URL (so equity is preserved); redirects and removals are sent to your site so it can honor them. It's [content consolidation](https://skribra.com/glossary/content-audit) done with a human in the loop.

## How it works

### Review monthly

Curator analyzes every published article for overlap and low value, using GSC to rank the keeper in each cluster.

### Recommend, don't act

It writes per-article suggestions — recreate, redirect, noindex, or remove — and surfaces them on your dashboard.

### You approve

You accept or dismiss each one. Approved actions are applied and, for redirects/removals, sent to your site to honor.

**Related feature:** [Fix keyword cannibalization](https://skribra.com/features/keyword-cannibalization)

**Documentation:** [How to support redirects & removals on your site](https://skribra.com/docs/integrations/custom-webhook#content-health)

## Machine-readable index

- HTML version of this page: https://skribra.com/agents/curator
- Site index for agents: https://skribra.com/llms.txt
- Sitemap: https://skribra.com/sitemap.xml
- Public REST API: https://skribra.com/api/v1
- OpenAPI specification: https://skribra.com/openapi.json
