Logly Releases "mureo," an Advertising Operations Framework for AI Agents, as Open Source
Logly, Inc. has released "mureo," a framework that allows AI agents to autonomously manage advertising accounts, as open-source software. It features cross-platform integration and robust security.
📋 Article Processing Timeline
- 📰 Published: April 23, 2026 at 00:30
- 🔍 Collected: April 23, 2026 at 00:02
- 🤖 AI Analyzed: April 23, 2026 at 03:17 (3h 15m after Collected)
Logly, Inc. (Headquarters: Shibuya-ku, Tokyo; Representative Director: Hirokazu Yoshinaga; Securities Code: 6579) announced that it has released "mureo," a framework for AI agents to autonomously manage advertising accounts, as open-source software under the Apache License 2.0.
Official Website: https://mureo.io
GitHub Repository: https://github.com/logly/mureo
Logo Image
Background of the Release
With the rapid evolution of generative AI, the use of AI agents (such as Claude Code, Cursor, Codex CLI, and Gemini CLI) is expanding in advertising operations. However, many conventional AI tools are limited to executing single instructions or retrieving reports, and fail to adequately support continuous operations that reflect business strategies and the judgment criteria of veteran operators. Furthermore, while cross-platform analysis and decision-making involving Google Ads, Meta Ads, Google Search Console, and Google Analytics 4 (GA4) are daily tasks in practice, support from AI tools in this area has been limited.
Based on our accumulated advertising operations know-how as an ad technology company and our internal experience with AI agents, Logly developed the framework "mureo" to solve these challenges. To address the industry issue of manpower shortages in ad operations and to contribute to the overall development of the industry by broadening the base of AI utilization, we decided to release it as open source.
Features of "mureo"
Decision-Making Starting from Business Strategy
mureo manages ads based on the business strategy set by the user (persona, USP, brand voice, KPI, operation mode). Rather than mere numerical optimization, it allows users to entrust business-aligned judgments to the AI agent, realizing operations consistent with corporate brand policies—from creating ad copy to proposing budget allocations.
Integrated Workflow Across Platforms
It processes Google Ads, Meta Ads, Google Search Console, and GA4 in a single workflow. It collectively retrieves ad delivery status/results, organic search trends, and user behavior on the site, automatically performing correlation analysis. This eliminates the manual matching of fragmented data by operators and enables multidimensional decision-making, such as identifying overlaps between paid ads and organic searches or performance differences by device. Integration with GA4 is achieved by combining an external MCP server.
Systematizing the Judgment Criteria of Veteran Operators
We incorporated into the workflow the judgment criteria that experienced operators have acquired over years of practice: identifying causes of non-delivery, pre-submission ad checks, organizing search terms, evaluating budget efficiency, analyzing device-specific performance differences, and verifying consistency with landing pages. Even teams trusting AI with ad operations for the first time can start with a quality level akin to having a veteran operator guiding them.
Operational Knowledge That Gets Smarter with Use
The agent learns operational insights such as "CPAs rise during consecutive holidays in this industry" or "Since our product is B2B, weekend numbers are not a reference," and automatically reflects them in subsequent sessions. It makes judgments understanding account-specific contexts, growing into a partner that aligns closer to your operation policies the more it is used.
Robust Security Design Specialized for AI Agent Operations
Entrusting ad operations to an AI agent introduces new risks distinct from conventional SaaS, such as credential leaks, unintended budget changes, and account suspensions due to ad disapprovals. Operating on these premises, mureo comes standard with a multi-layered defense (Defense-in-Depth) architecture not seen in typical SaaS.
By layering multiple defenses—strict isolation of credentials, blocking of malicious queries, real-time anomaly detection based on statistical baselines, full recording of write operations with single-command rollbacks limited to safe scopes, and pre-submission policy checks—it ensures that accidental operations or prompt injections do not affect the advertising account. Furthermore, through a fully local architecture with absolutely no telemetry or external transmission functions, the path for credentials and operational data to leak to third parties is fundamentally eliminated. While delegating operations to AI, humans retain final decision-making and responsibility, making it a framework designed for companies to safely entrust their advertising accounts.
Main Workflow Commands
- /onboard
Official Website: https://mureo.io
GitHub Repository: https://github.com/logly/mureo
Logo Image
Background of the Release
With the rapid evolution of generative AI, the use of AI agents (such as Claude Code, Cursor, Codex CLI, and Gemini CLI) is expanding in advertising operations. However, many conventional AI tools are limited to executing single instructions or retrieving reports, and fail to adequately support continuous operations that reflect business strategies and the judgment criteria of veteran operators. Furthermore, while cross-platform analysis and decision-making involving Google Ads, Meta Ads, Google Search Console, and Google Analytics 4 (GA4) are daily tasks in practice, support from AI tools in this area has been limited.
Based on our accumulated advertising operations know-how as an ad technology company and our internal experience with AI agents, Logly developed the framework "mureo" to solve these challenges. To address the industry issue of manpower shortages in ad operations and to contribute to the overall development of the industry by broadening the base of AI utilization, we decided to release it as open source.
Features of "mureo"
Decision-Making Starting from Business Strategy
mureo manages ads based on the business strategy set by the user (persona, USP, brand voice, KPI, operation mode). Rather than mere numerical optimization, it allows users to entrust business-aligned judgments to the AI agent, realizing operations consistent with corporate brand policies—from creating ad copy to proposing budget allocations.
Integrated Workflow Across Platforms
It processes Google Ads, Meta Ads, Google Search Console, and GA4 in a single workflow. It collectively retrieves ad delivery status/results, organic search trends, and user behavior on the site, automatically performing correlation analysis. This eliminates the manual matching of fragmented data by operators and enables multidimensional decision-making, such as identifying overlaps between paid ads and organic searches or performance differences by device. Integration with GA4 is achieved by combining an external MCP server.
Systematizing the Judgment Criteria of Veteran Operators
We incorporated into the workflow the judgment criteria that experienced operators have acquired over years of practice: identifying causes of non-delivery, pre-submission ad checks, organizing search terms, evaluating budget efficiency, analyzing device-specific performance differences, and verifying consistency with landing pages. Even teams trusting AI with ad operations for the first time can start with a quality level akin to having a veteran operator guiding them.
Operational Knowledge That Gets Smarter with Use
The agent learns operational insights such as "CPAs rise during consecutive holidays in this industry" or "Since our product is B2B, weekend numbers are not a reference," and automatically reflects them in subsequent sessions. It makes judgments understanding account-specific contexts, growing into a partner that aligns closer to your operation policies the more it is used.
Robust Security Design Specialized for AI Agent Operations
Entrusting ad operations to an AI agent introduces new risks distinct from conventional SaaS, such as credential leaks, unintended budget changes, and account suspensions due to ad disapprovals. Operating on these premises, mureo comes standard with a multi-layered defense (Defense-in-Depth) architecture not seen in typical SaaS.
By layering multiple defenses—strict isolation of credentials, blocking of malicious queries, real-time anomaly detection based on statistical baselines, full recording of write operations with single-command rollbacks limited to safe scopes, and pre-submission policy checks—it ensures that accidental operations or prompt injections do not affect the advertising account. Furthermore, through a fully local architecture with absolutely no telemetry or external transmission functions, the path for credentials and operational data to leak to third parties is fundamentally eliminated. While delegating operations to AI, humans retain final decision-making and responsibility, making it a framework designed for companies to safely entrust their advertising accounts.
Main Workflow Commands
- /onboard