How logly/mureo works
mureo is a local-first control plane for AI-assisted advertising operations. An agent uses its MCP tools to inspect account data, analyze performance against strategy, and—when connected to live APIs—make safety-gated changes; outcomes and tool calls are recorded for review and rollback. Users can connect live platforms or analyze imported BYOD data, and can configure the local installation through a browser UI. The graph includes documented capabilities, but sampled excerpts are incomplete, so some subsystem connections are omitted where their implementation path is not evidenced.
Main entry points: server.py MCP Server · client_factory.py Client Factory · registry.py Analytics · tools_google_ads.py Platform Tools
Components 20
Agent Access
- MCP Server
mureo/mcp/server.py - Platform Tools
mureo/mcp/tools_google_ads.py
Data Connections
- Client Factory
mureo/client_factory.py - Google Ads
mureo/google_ads/client.py - Meta Ads
mureo/meta_ads/client.py - BYOD Clients
mureo/byod/clients.py - Amazon Ads Bridge
mureo/amazon_ads/provider.py - Search Console
mureo/search_console/client.py - Credentials
mureo/auth.py
Strategy and Analysis
- Analytics
mureo/analytics/registry.py - Delivery Analysis
mureo/analysis/delivery_collapse.py - Strategy and State
mureo/context/state.py - Safety Policy
mureo/policy/guardrails.py - Account Learning
mureo/learning/context_builder.py
Auditable Operations
- Action Log
mureo/context/action_log_store.py - Tool Journal
mureo/mcp/journal.py - Rollback
mureo/rollback/executor.py
Local Setup
- Configure UI
mureo/web/server.py - Setup Routes
mureo/web/handlers.py
Other components
- AI Agent
Connections 21
- AI Agent → MCP Server: invokes tools
- MCP Server → Platform Tools: exposes tools
mureo/mcp/server.py - Platform Tools → Client Factory: opens clients
mureo/client_factory.py - Client Factory → Google Ads: creates client
mureo/client_factory.py - Client Factory → Meta Ads: creates client
mureo/client_factory.py - Client Factory → BYOD Clients: routes BYOD
mureo/client_factory.py - AI Agent → Analytics: requests analysis
- Analytics → Client Factory: opens live clients
- Analytics → Delivery Analysis: runs analysis
- Delivery Analysis → Strategy and State: reads strategy
mureo/analysis/delivery_collapse.py - MCP Server → Strategy and State: serves context
- AI Agent → Safety Policy: follows safety gates
- AI Agent → Action Log: records actions
- MCP Server → Tool Journal: journals tool calls
mureo/mcp/server.py - AI Agent → Rollback: requests rollback
- AI Agent → Account Learning: builds knowledge
- Configure UI → Setup Routes: dispatches routes
mureo/web/server.py - AI Agent → Configure UI: configures locally
- Setup Routes → Credentials: sets up credentials
mureo/web/handlers.py - MCP Server → Amazon Ads Bridge: bridges provider
mureo/mcp/server.py - AI Agent → Search Console: uses Search Console