Thesis Tutor MCP Server

What it solves

Common thesis blockers are not simply lack of writing ability; they include unstable topic choice, repeated outline changes, hard-to-apply advisor feedback, missing pre-submission checks, and limited rehearsal for the defense. Thesis Tutor MCP packages the thesis tutoring workflow into a set of MCP tools that AI agents can call, making it useful when thesis assistance needs to be embedded in an existing workflow, RAG assistant, or agent orchestration layer.

How it works

It wraps the thesis-tutor skill as an MCP server and converts thesis tasks into tool calls:

• Topic phase: get_topic_suggestions generates candidate topics, supported by the built-in topic_selection guide.
• Writing phase: generate_outline builds an outline, review_chapter annotates a paragraph, and interpret_feedback turns advisor feedback into actionable revisions.
• Quality checks: run_checklist runs pre-submission checks, while plagiation_guide covers plagiarism and AI-detection considerations.
• Defense and materials: simulate_defense rehearses the defense, and list_references / get_reference query built-in reference files.

Callers do not need to know the internal templates; they select tools by task, and MCP returns structured advice or document content.

Caveats

The skill is oriented toward tutoring and checks, not a replacement for advisor review, university plagiarism systems, or formal academic standards. Its built-in materials focus on method and writing support, such as literature reviews, quantitative/qualitative research, and outline templates, so they should not be used directly as factual thesis citations. Treat outputs as drafts, revision lists, and practice materials; verify key claims against discipline requirements and primary sources.