--- title: "Google Search Console Integration" description: "Connect AI agents to Google Search Console. Agents query search performance by page, query, country, and device, submit and audit sitemaps, inspect URL index status and canonicals, and manage properties across accounts." url: https://flowrunner.ai/integrations/google-search-console date_modified: 2026-08-04T10:04:16-07:00 --- # Google Search Console [SEO](https://flowrunner.ai/integrations/category/marketing-social) Connect AI agents to Google Search Console. Agents query search performance by page, query, country, and device, submit and audit sitemaps, inspect URL index status and canonicals, and manage properties across accounts. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 10 actions · OAuth · available [Google Search Console website](https://search.google.com/) · [Platform Documentation](https://developers.google.com/webmaster-tools) · Capability data verified 2026-08-12 1. A weekly schedule fires Monday morning, after the data delay window closes 2. Query Search Analytics pulls the prior week by query and page, and a second call pulls the same window from the week before 3. The agent diffs the two windows and flags pages where clicks or average position moved beyond the set bands 4. For each flagged page, Inspect URL returns the current index status, last crawl, and the Google-selected canonical 5. The report posts to the SEO channel: movers, index verdicts, and canonical mismatches, with sheet rows appended for trend history 6. The content lead reads the flagged set and decides what gets rewritten, redirected, or left alone 7. Decisions log back to the content backlog with the evidence attached ## What This Integration Enables Search Console is where Google tells you what it thinks of your site, and almost every team reads it the same way: occasionally, manually, and after the damage. FlowRunner agents read it on a discipline. Query Search Analytics exposes clicks, impressions, CTR, and average position across up to 25,000 rows per request, grouped by date, query, page, country, device, or search appearance, which is enough resolution to treat organic performance as a monitored system instead of a monthly anecdote. Around it, the sitemap and inspection operations turn "did Google index our stuff" from an assumption into a verifiable pipeline step. - Pull search performance by query, page, country, and device into sheets, dashboards, and weekly diffs - Alert on position drops and click decay the week they happen, with the affected queries attached - Submit sitemaps after publishes and audit error, warning, and indexed-URL counts on schedule - Verify individual URLs with Inspect URL, including index verdict and Google-selected versus declared canonical - Manage and audit properties across accounts, with destructive property operations gated ### Without FlowRunner **Reporting by screenshot**: Someone opens Search Console, eyeballs the graph, and pastes a screenshot into the Monday deck **Decay noticed at quarter end**: A page that slid from position 3 to position 9 in March becomes visible in the June traffic review **Indexing on faith**: Pages ship, sitemaps sit, and whether Google actually indexed the new section is nobody's explicit job ### With FlowRunner **The diff arrives, not the dashboard**: Agents pull the same windows every week and report what changed beyond thresholds, with history accumulating in sheets **Decay caught within its first week**: Position drops surface Monday with the affected queries attached, while the fix is still cheap **Publish-to-indexed as a checked pipeline**: New URLs get sitemap resubmission and Inspect URL verification, and a human hears about every FAIL verdict ## Use Case Scenarios ### The Monday diff that replaced the Monday screenshot An agent pulls last week and the week prior from Query Search Analytics, grouped by page and query, and appends both to a trend sheet in [Google Sheets](https://flowrunner.ai/integrations/google-sheets). Pages that lost more than the threshold in clicks or slipped in average position post to the SEO channel in [Slack](https://flowrunner.ai/integrations/slack) with their top affected queries. The content lead spends Monday deciding, not collecting. When a money page slips, the rewrite starts that week. ### Publish, submit, verify, and know The content pipeline ships a batch of new pages. The workflow calls Submit Sitemap to request reprocessing, waits out the crawl delay, then runs Inspect URL against each new URL. PASS verdicts log quietly. A FAIL or a canonical pointing somewhere unexpected posts to the channel with the coverage state and last crawl time, and the fix happens before anyone builds links to a page Google refuses to index. ### One report across every property An agency manages a dozen client properties. List Sites enumerates them with permission levels, and the weekly agent loops Query Search Analytics across each, normalizing the output into one cross-client dataset. Per-client tabs feed the reporting decks; an exceptions view surfaces any property where sitemap errors climbed or a permission unexpectedly downgraded to restricted. The account team reviews exceptions, not twelve dashboards. ## Human-in-Loop Highlight Delete Site looks administrative and acts structural. It unlinks a property from the connected account, and every workflow authenticating through that account loses the property with it: the weekly analytics pulls, the sitemap audits, the URL inspections, all of them start failing at once, and if the connected user was the only one holding access, re-verification stands between you and your own historical reporting pipeline. Delete Sitemap has the same trap in miniature, removing a submission that another team relied on Google reprocessing. Neither operation touches the website itself, which is exactly why an agent can execute them without visible damage until the next scheduled run fails. So FlowRunner gates property-level removals: the agent states which property, which workflows currently depend on it, and what stops working, and a person with owner context confirms. Reads run free. Removing the thing every read depends on requires a human who knows why it was there. 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 10 actions ### Search Performance 1 - **Query Search Analytics** Queries clicks, impressions, CTR, and average position for a property, grouped by up to several dimensions including Date, Query, Page, Country, Device, and Search Appearance, with filtering by search type and dimension filters. Returns up to 25,000 rows per request with Start Row pagination. Dates are YYYY-MM-DD in PT, and data typically trails by 2 to 3 days unless the fresh Data State is requested. ### Properties 4 - **List Sites** Lists all properties accessible to the connected user, verified and unverified, with each site URL and the user's permission level. The enumeration step for multi-property reporting and permission audits. - **Get Site** Retrieves a single property with its site URL and the user's permission level. Fails if the property is not in the user's account. - **Add Site** Adds a property to the connected user's account. Ownership still has to be verified through Search Console before most data becomes available. - **Delete Site** Removes a property from the connected user's account. It does not delete Search Console data or affect the website, but every workflow reading through this account loses the property. Gated by this page's human check. ### Sitemaps 4 - **List Sitemaps** Lists sitemaps submitted for a property, or those referenced by a sitemap index, with type, submission and last-download timestamps, pending status, error and warning counts, and submitted versus indexed URL counts. - **Get Sitemap** Retrieves one sitemap's processing detail: whether Google has downloaded it, its errors and warnings, and how many of its URLs were indexed. - **Submit Sitemap** Submits a sitemap for crawling and processing; resubmitting an existing sitemap requests reprocessing. The file must already be live on the site. Paired with Get Sitemap to confirm processing. - **Delete Sitemap** Deletes a sitemap submission from the property. The file stays on the website and Google may still crawl known URLs, but the submission record and its processing status are gone. ### URL Inspection 1 - **Inspect URL** Inspects a URL and returns Google's index status: PASS, FAIL, or NEUTRAL verdict, coverage state, last crawl time, robots.txt and indexing allowance, and the Google-selected versus user-declared canonical, plus mobile usability and rich results where available. Rate-limited by Google to about 2,000 inspections per property per day, so batch workflows budget it deliberately. ## Frequently Asked Questions ### What can FlowRunner do with Google Search Console? FlowRunner agents can run Query Search Analytics, List Sites, and Get Site in Google Search Console, plus 7 more actions. ### Does connecting Google Search Console to FlowRunner require OAuth? Yes. Google Search Console connects to FlowRunner with OAuth 2.0, so agents authenticate without handling raw credentials. ### Can Google Search Console trigger a FlowRunner workflow automatically? Google Search Console doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at Google Search Console?** This integration exposes Google Search Console 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/google-search-console. Site index: https://flowrunner.ai/llms.txt