mcp-persona-sessions

What is mcp-persona-sessions?

mcp-persona-sessions is a Model Context Protocol (MCP) server that lets AI assistants run structured, persona-driven sessions such as mock interviews, personal reflection, and coaching conversations. It is designed for users who want guided dialogue with realistic personas in a safe environment.

How to use mcp-persona-sessions?

Clone the repository, install the fastmcp dependency, copy config.yaml.example to config.yaml and edit it, then add the server definition to your MCP client configuration (e.g., Claude Desktop). To start a session, ask the AI assistant to initiate one with a specific persona and topic; the assistant will use the available tools to manage the conversation.

Key features of mcp-persona-sessions

• Load detailed persona profiles that transform AI behavior and expertise.
• Track session duration with start, stop, and status checking.
• Pre-defined session templates with clear goals and outcomes.
• Adaptive session flow that responds to natural conversation.
• Detailed feedback on performance and communication effectiveness.
• Secure file handling with path validation and error handling.

Use cases of mcp-persona-sessions

• Practice job interviews with realistic interviewer personas.
• Rehearse presentations and get feedback on messaging.
• Role-play difficult workplace conversations safely.
• Engage in structured self-reflection and goal clarification.
• Conduct guided sessions for creative exploration and decision-making.

FAQ from mcp-persona-sessions

What prerequisites are needed to run the server?

Python 3.8 or later and an MCP-compatible AI assistant (such as Claude Desktop) are required.

How do I add a custom persona?

Create a new Markdown file in the roles/ directory with sections for Identity and Purpose, Communication Style, Expertise Areas, and Goals.

How do I add a custom session type?

Edit the session_types.yaml file and add a new entry with fields such as name, persona, description, duration, topics, and goals.

Can I get feedback on a session after it ends?

Yes, the get_session_feedback tool analyzes the session transcript and provides detailed evaluation.

Where are persona definitions stored?

Persona files are stored in the roles/ directory, and the path is configured in config.yaml under the persona_path setting.