タラ・マイヤー
7月 16, 2026
New functionality lets growth teams create, update, and manage campaigns, apps, callbacks, and fraud filters directly through AI assistants, no switching between tools required.
Tenjin has announced the launch of write capabilities for its Model Context Protocol (MCP) Server, making it the first mobile measurement partner (MMP) to enable AI assistants to take action not just answer questions within a marketer’s Tenjin account.
The announcement extends Tenjin’s existing MCP Server, which already allows AI assistants to query live attribution, monetization, and performance data. With today’s update, AI assistants can now create, update, and delete campaigns, apps, S2S callbacks, and fraud filters entirely through natural language, without requiring users to leave their AI workspace.
Eliminating the Last Mile of Fragmentation
Tenjin was founded on the premise that fragmented data slows growth teams down. Today, the company is applying that same philosophy to create a truly integrated workflow instead of tool fragmentation. In particular, the constant context-switching between AI assistants and marketing platforms that interrupts analyst and UA manager productivity.
“Mobile marketers are already using AI assistants every day, but the moment they need to act on what they’ve found, they have to leave the conversation and log into another tool,” said Roman Garbar, Marketing Director at Tenjin. “We built write capabilities into our MCP Server so that querying data and acting on it can happen in the same place, in plain language.”
Key Capabilities
Tenjin’s MCP Server now supports a broad range of read and write operations, including:
- Campaign management: Create, update, delete, and search campaigns across ad networks
- App management: Add and configure apps, and run SDK integration health checks
- Callback management: Build and update S2S callbacks, callback settings, and callback groups
- Fraud prevention: Block, restore, review, or clear Site ID filters for traffic quality control
- Reporting and analytics: Run attribution, monetization, SKAN, pLTV, and ad spend reports with flexible filtering and cohort-aware logic
- Dry runs: Preview any create, update, or delete action before changes are applied
- Bulk operations: Execute multiple updates in a single request with per-item success reporting
All capabilities are easily accessible through conversational prompts such as “Create a new AppLovin Android campaign” や “Block these Site IDs” so no API knowledge or technical background is required.
Hosted Infrastructure and Zero-Configuration Setup
Unlike many MCP implementations that require local installation or developer configuration, Tenjin’s MCP Server is fully hosted. It is accessible across platforms including the Claude and Codex web and desktop apps and is available through each platform’s plugin marketplace with no terminal setup needed.
This also means that updates to tools and Skills are delivered automatically, ensuring users always have access to the latest capabilities without manual maintenance.
Skills: Teaching AI Agents Workflows
As mentioned earlier, Tenjin ships Skills curated guidance that teaches AI assistants the conventions of user acquisition and common Tenjin workflows, from campaign setup to SDK integration and attribution configuration. Where tools define what an AI can do, Skills inform how and when to apply them, resulting in more accurate, reliable multi-step workflows out of the box.
How to Get Started with Tenjin’s MCP Server
Users can get started by installing the Tenjin plugin through their AI assistant’s plugin marketplace and logging in with an existing Tenjin account. Full setup instructions are available at Tenjin’s Getting Started guide.
For more information, contact support-jp@tenjin.com. We’re happy to help you get connected.
