PRO Annual — 50 users, $106.80/year
Stop guessing why content fails to rank. When Claude runs AddSearch via an MCP, you instantly audit your site's search performance, identifying critical content gaps and zero-result queries directly inside GitScrum.
Claude
ChatGPT
Cursor
Gemini
Windsurf
VS Code
JetBrains
VercelConnected via Vinkius catalog · No credit card
task.delegate()
You start the process by asking Claude: 'Check our top 5 product categories for zero-result search queries last week, and list any required content additions.'
Claude automatically invokes the AddSearch MCP tools. It executes advanced filtering on your index to gather live analytics data points.
The resulting list of content gaps, along with suggested new tickets for the backlog, appears directly on your sprint board for immediate assignment and review.
task.create()
Copy. Paste into a Claude-aware GitScrum task. Done.
A Content Manager realizes they can't manually check if a new article is being found by users searching different keywords.
I need to audit the content gaps for our e-commerce site. Using AddSearch, please pull all zero-result search queries from the last 30 days and organize them into a list of actionable tickets. For each query, suggest the appropriate category and product line.
A developer knows a specific URL's metadata was corrupted or needs to be updated in the search index immediately.
I found that product page 'X' has incorrect inventory data on its site index. Using AddSearch, please update the document for this URL with the following JSON payload and confirm the indexing process is complete.
A PM needs to ensure that when a user searches for 'red shoes', the search results correctly prioritize products from the Nike brand only.
Using AddSearch, please execute a filtered search for 'shoes' where the custom field 'color' equals 'red' AND the custom field 'brand' equals 'nike'. Then, list the top 10 results and confirm they meet our ranking criteria.
Manually checking site search performance is a nightmare. You spend hours hopping between Google Search Console, analytics dashboards, and Slack threads just to compile a list of content gaps or identify poor keyword coverage. It's slow, prone to copy-paste errors, and often outdated by the time you finish.
The new flow changes everything. Claude reads your live search index data via this AddSearch MCP connection and writes the outcome directly into GitScrum. You instantly get a structured list of actionable tickets detailing exactly what content needs writing or updating.
Open this workflow in GitScrum →Debugging the search index used to involve manual API calls, complex spreadsheet lookups, and waiting for multiple team members to confirm a document was properly ingested. This process is slow and breaks development flow every time.
Now, Claude handles the heavy lifting. Using AddSearch MCP in your preferred AI client like Cursor or VS Code, it can list all indexed documents or push fixes with a single prompt. Your dev sprint stays focused on building, not debugging search plumbing.
claude.ask()
I need to check if any of our key product lines are suffering from poor search visibility. Can Claude use the AddSearch MCP to run a query across all categories and report back on potential weak spots?
Can Claude analyze our last week's traffic data using AddSearch? I want to see what users searched for that didn't lead to any results, so we can prioritize content creation.
I just uploaded 20 new product descriptions. Before they go live, can Claude use AddSearch MCP to index them and then check the metadata structure against our standard template?
Claude, please simulate a user experience by running an autosuggest test on our main search bar using these five keywords: [list of keywords].
tools.list()
Each tool is one operation Claude runs on your behalf.
| tool_name | operation |
|---|---|
| delete_document | Claude can permanently delete a document from the search index, useful for removing outdated content or sensitive information. |
| search_filtered | The AI agent can perform targeted searches within your indexed content by applying specific custom field filters, such as searching only products in 'shoes'. |
| index_document | Requires Secret Key. Claude can add a brand new document to the index, or update an existing one, keeping your search data fresh. |
| list_documents | Requires Secret Key. The agent lists every single indexed document name and its basic metadata for a full content audit overview. |
| search_pagination | Claude retrieves specific pages of search results, allowing you to test how the search index handles large volumes of data like real user browsing. |
| search_keyword | The AI agent searches your entire indexed content using one or more general keywords. This is useful for broad topic research in a sprint. |
| search_sorted | Claude performs advanced searches and ensures the results are ordered by a custom variable, helping you verify search ranking rules. |
| stats_clicks | Requires Secret Key. The agent retrieves click-through analytics, showing which search result links users actually clicked on after searching. |
| stats_queries | Requires Secret Key. Claude pulls comprehensive search query data, identifying the most common searches and any major zero-result gaps for content planning. |
| autosuggest | The AI agent gets autocomplete suggestions based on a prefix keyword. This is perfect for testing user expectation and improving front-end UX. |
grade
A+
tools
10
auth
Required
catalog
Vinkius
faq.read()
Free for small teams. Set up in 5 minutes. Cancel anytime.
GitScrum lists this MCP as a Vinkius partner. Installations happen on Vinkius.
Connect GitScrum with the tools your team already uses. Native integrations with Git providers and communication platforms.
GitHub
GitLab
Bitbucket
Slack
Teams
Discord
Zapier
PabblyConnect with 3,000+ apps via Zapier & Pabbly
