Semly MCP Apps - capabilities, permissions and security

In simple terms, MCP is a secure bridge between Semly and your AI assistant. It allows ChatGPT, Claude, Cursor, Codex, and other compatible tools to work with Semly without you manually pasting reports into the conversation. Within the permissions you grant, the assistant can retrieve the right data, analyze it, and take the next steps.

Instead of exporting a report, copying data into AI, and implementing recommendations separately, you can give one instruction: “Find out why my visibility dropped, compare me with competitors, and recommend what to do next.” The assistant can retrieve the required Semly data and move from analysis to action.

What changes in practice?

After connecting MCP, you are not using AI only for conversation. You are giving it controlled access to Semly data and functions. This allows one task to span several areas without manually moving data between tools.

• check brand visibility across AI models and compare it with competitors

• read actual model answers and inspect the sources they use

• combine AI visibility data with Google Search Console and Bing Webmaster Tools

• find content gaps, lost prompts, cannibalization, and indexing issues

• suggest new monitoring prompts and add them after approval

• prepare an article brief from real data, generate the content, and edit it

• publish content to a connected channel after your approval

Why is this so powerful?

The key value of MCP is that it connects analysis and execution in one workflow. The assistant can identify a problem from real data, investigate its cause, recommend a solution, and carry out an approved action.

Example: “Find prompts where we lost visibility during the last 30 days. Check which competitors replaced us, which sources are cited in current answers, compare this with GSC and Bing, then suggest 3 article topics. Prepare a brief for the best one for my approval.”

After approval, you can continue in the same conversation:

“I approve the brief. Generate the article, improve the title and meta description, then publish it to the selected channel.”

This means AI is not limited to answering questions about your data. It can become an operational assistant for improving brand visibility.

Semly as the data and action layer for AI

Semly can provide the model with the wider brand context, including visibility data, prompts, answers, sources, competitors, recommendations, Google Search Console, Bing, articles, and configured publishing channels. One assistant can therefore work from a shared view of the situation instead of analyzing every system separately.

MCP is available on Ultra, Scale, and Enterprise plans. It is not available on Premium.

Where to find MCP

• Settings -> MCP Apps

• Integrations -> MCP Apps

https://api.semly.ai/mcp

What can an AI assistant analyze?

• brand visibility and trends

• Share of Voice, average position, and AI Growth Score

• prompt and topic statistics

• full AI model answers

• sources and domains cited by models

• competitor visibility and content gaps

• Semly reports

• Google Search Console

• Bing Webmaster Tools and Google vs Bing comparisons

• Semly recommendations

• content statistics and bot visits

Google Search Console and Bing

When these integrations are connected to a brand, the model can analyze pages, queries, clicks, impressions, CTR, positions, cannibalization, index status, and Google vs Bing comparisons.

Prompts, brands, and recommendations

With the right permissions, an assistant can add prompts individually or in batches, create brands, and update recommendation statuses.

Articles and publishing

• generate an article

• prepare a brief for approval before generation

• edit title, slug, meta description, and HTML content

• publish to the Semly Knowledge Base

• publish to WordPress

• publish to Shoper

• publish to Shopify

• publish to RedCart

• publish to Shoptet

Permissions

• Read

• Update recommendation statuses

• Create prompts

• Create brands

• Generate articles

• Edit articles

• Publish articles

Leon and token optimization

Semly uses its AI Agent Leon as an intermediary layer. The model can agree with Leon on the exact scope of data it needs instead of retrieving entire reports. This helps reduce token usage and keeps the analysis focused.

Logs and security

Semly lets you review MCP operation history and revoke a connection at any time. For analysis-only use cases, Read permission is usually enough.

Example

Analyze the last 90 days of Semly, Google Search Console, and Bing data. Find the largest content gap versus competitors, inspect AI answers and sources, prepare an article brief for my approval, then generate and publish the article after approval.