If you use an AI client for research, you need a way to access OppAlerts records within that process. OppAlerts provides a customer-facing MCP server. It exposes every read operation available through the API, including running searches and retrieving the customer's stored data and reports. It cannot create or edit campaigns or change configuration.
Compatible AI tools connect through OAuth. MCP access is included; published request limits are still being finalized.
What MCP is
MCP, the Model Context Protocol, is an open protocol that lets an AI application request the tools and data a connected server exposes. The protocol documentation is the reference for how it works.
The practical effect for a marketer: instead of exporting a report and pasting it into an assistant, a compatible application can request the data directly while answering your question.
How MCP supports research
The reports in this product are research inputs rather than final answers. Someone reads a prospect list and decides which rows to pursue, or reads tracked answers and decides what to write.
That is the type of work where having the data available inside an assistant changes the workflow rather than saving a copy and paste. Asking which of two hundred prospects share a characteristic is a question a person answers slowly and an assistant answers quickly, given access.
What to establish before it influences a purchase
- Which tools the server exposes. A server offering one report is different from one offering the research functions.
- Which applications are supported, since client compatibility varies and changes.
- How access is authenticated, and whether a client's data can be isolated.
- What the request allowance is and whether it is counted against your collection allowance.
The protocol being open does not answer any of these. They are decisions each vendor makes.
Where competitors stand today
Several products already provide MCP access: Ahrefs on every plan, Semrush from Starter, Serpstat with ChatGPT and Claude integration, Brand24, Surfer, AccuRanker, and Otterly.AI. Otterly publishes monthly MCP request allowances.
OppAlerts also meets a read-only MCP requirement today. The server is an interface over the same API, so its available data operations match the API exactly.
When preparing the connection, choose a supported read operation and retain the source evidence needed to check the client’s research output.
See the API or the current status.