DVC MCP Server

Connect your DVC Studio account to any AI agent and take full control of your machine learning experiments and data versioning workflows through natural conversation.

What you can do

• Project Orchestration — Expose registered organization workspaces and validate available physical repositories connected within DVC Studio limits
• Experiment Navigation — Iterate through explicitly generated model runs mapping precise metric arrays and discovering logged metrics history cleanly
• View Management — Extract explicit UI configuration layouts and dashboard settings to retrieve structural workspace representations natively
• Repository Auditing — Analyze specific identifier boundaries resolving internal team mappings and parsing direct repository metadata constraints
• Metric Inspection — Retrieve complex structural arrays defining precisely which metrics were captured during specific experiment epochs
• Identity Oversight — Identify the exact authorized token holder exposing mapping roles and organization scopes dynamically to verify permissions

How it works

1. Subscribe to this server
2. Enter your DVC Studio Client Access Token (found in DVC Studio Profile Settings > Tokens)
3. Start managing your ML experiments from Claude, Cursor, or any MCP-compatible client

Who is this for?

• Data Scientists — monitor model experiments and track metrics without leaving the development environment
• ML Engineers — verify project repository connections and audit model run histories using natural language
• Team Leads — monitor organization workspaces and experiment progress through natural conversation
• DevOps Engineers — test and debug DVC Studio integrations and verify access token permissions in real-time