Skip to main content

Fix what the crawler sees first.

Tracks Google positions from Search Console run over run and audits your canonical pages with 19 checks. A slipping page lands in your queue as one ranked fix.

Free analysis · 5 credits · no card · first score in about 30 seconds

Google Search Console · 19 checks · up to 21 pages per audit

Fix the missing canonical on /pricingIn your queue
check canonical · FAIL
observed → /pricing-old → 301 → /pricing
expected canonical href → …/pricing
impact +2 score · indexed duplicates removed

sample run · demo data

01 / The problem it removes

Rankings decay quietly.

A page slips for weeks before traffic drops. The cause, a missing canonical or a stale title, sits in tools you do not open.

Without it

  • The position drops weeks before the traffic graph shows it.
  • The cause hides in tools you never open.
  • Monitoring and fixing live in different places, so the fix ships late.

With the SEO agent

  • Positions are compared run over run, before traffic drops.
  • 19 checks run on every sampled page, observed value against expected.
  • The failed check and its fix arrive in the same ranked queue.

02 / How it runs

A loop, not a task list.

  1. 01

    Crawl

    Reads the pages that carry your traffic and records their technical state.

  2. 02

    Track

    Your Google positions from Search Console are compared against the previous run.

  3. 03

    Flag

    A slipping page or failed check becomes one ranked fix with evidence.

  4. 04

    Re-check

    The next scheduled run measures whether the repair worked.

03 / What it can do

What the agent does

Search tracking

Google positions from Search Console, compared run over run.

Technical audit

Canonicals, indexing, sitemaps, structured data, llms.txt.

Page-level evidence

Failing URL, observed value and expected value, side by side.

Early warnings

Two adjacent 28-day windows, weighted by impressions.

Ranked fixes

Failed check to next action, no second tool.

Scheduled re-checks

Every repair is measured on the next run.

04 / In the product

A recommendation lands. You ship it in minutes.

Every SEO agent item arrives with evidence and a ready prompt. Copy it into Claude Code or Cursor, review the diff, approve. The next run measures whether it moved the score.

didit.me · real run · Aug 27, 2026
The site audit for didit.me: the technical check rows with their pass and fail state

Review queue

In your queue
Fix the missing canonical on /pricing
↓ then paste it
your terminal
claude "Fix the missing canonical on /pricing"
⏺ working… reading evidence from ai-cmo (MCP)
⏺ diff ready · 2 files changed
✓ shipped · re-measure queued for the next run

sample run · demo data

check canonical · FAIL observed → /pricing-old → 301 → /pricing expected canonical href → …/pricing impact +2 score · indexed duplicates removed

05 / Questions

Before you install it.

Two things, on the cadence you set: your Google positions from Search Console against the previous run, and a 19-check audit of up to 21 canonical pages picked from your home page, sitemap and Search Console. Each check is pass, fail or unknown, with evidence.

Your domain is enough for the technical audit. Search Console and Google Analytics are optional connections that add the search-visibility and traffic parts of the score. A part with no data is excluded rather than scored as zero.

Insights compare two adjacent 28-day windows: previous position minus current, weighted by that query's share of impressions. Ordering is deterministic, a recommendation is refused without traceable evidence, and the queue puts higher impact and confidence first.

No. Every failed check and slipping page arrives as a proposal you accept, decline or archive. It runs monthly on Free, weekly on Lite, daily on Pro, and switching it off changes nothing about the other agents.

Included in the free analysis

Start with the free analysis.

Free analysis, 5 credits that never expire, no credit card. The SEO agent briefs itself on your site during onboarding.