Tue-StudyOS

Build a deep in-session student research profile through an interview about interests, taste, skills, constraints, and thesis preferences. Use when asked to understand a student before generating research proposals, infer a thesis profile, prepare matching inputs, or turn course/interests text into a structured profile without storing private data.

Design, review, or improve portable Agent Skills with SKILL.md files. Use when asked to create a new skill, split an agent workflow into skills, write skill instructions, evaluate skill portability across agents such as Codex, Claude Code, OpenCode, or similar clients, or generate a skill package for thesis-finder workflows.

Draft concise, high-signal first-contact messages to potential thesis advisors using a deep student profile, a concrete research-proposal sketch, relevant papers, and chair fit evidence. Use when asked to write or improve an email to a professor, PhD student, postdoc, lab, chair, or thesis supervisor.

Find, rank, and summarize recent research papers for thesis discovery. Use when a student or coding agent asks for papers, publications, related work, recent research directions, literature leads, or thesis ideas based on interests, courses, skills, chairs, supervisors, or research areas.

Find university teaching chairs, research groups, professors, and possible thesis supervisors only after a sufficiently deep student research profile exists. Use when asked which chair, lab, professor, PhD student, research group, or university supervisor fits a topic, course profile, paper list, thesis direction, or research proposal.

Generate precise research-proposal sketches, thesis questions, and advisor-ready conversation starters from a deep student profile, recent paper evidence, and chair or supervisor matches. Use when asked for research proposals, thesis proposal ideas, proposal directions, topic hypotheses, or questions to discuss with a potential advisor.

Match a deep student research profile to professors, chairs, researchers, and recent papers using the professor seed index plus native web research. Use when asked to find proposal-relevant advisors, rank chairs or supervisors, compare matches, explain fit, or prepare evidence for precise research proposals without a database or UI.

Optionally maintain secondary thesis-finder Markdown data by fetching recent OpenAlex works for reviewed researchers and regenerating reviewable paper indexes. Use when running GitHub Actions, monthly maintenance, or manual data refresh workflows; not for student-facing thesis advice.

Discover and optionally install reusable agent skills from the open skills ecosystem. Use when the user asks "how do I do X", "find a skill for X", "is there a skill that can...", wants new capabilities, asks about installable skills, or before creating a new custom skill when an existing package may already cover the workflow.

Review StudyOS code changes for security, privacy, reliability, dead code, duplicate logic, regressions, missing validation, and maintainability. Use for PR reviews, pre-merge checks, "review this", "security review", "quality review", cleanup audits, or before claiming substantial StudyOS implementation work is ready.

Preserve durable StudyOS agent learnings as memory or reusable Codex skills. Use after substantial StudyOS work, repeated workflows, user corrections, non-obvious tool errors, runtime/deployment discoveries, or explicit requests to remember, self-improve, create skills, update skills, or refine agent procedures.

Create, update, or refine StudyOS Codex skills for recurring workflows. Use when the user asks for a new skill, skill creator flow, self-expansion, reusable workflow capture, procedure hardening, or when a completed task reveals a repeatable StudyOS workflow that should become or patch a skill.

Use when working in the tue-api-wrapper repository: adding university system endpoints, improving iOS/Electron/Next.js/ChatGPT surfaces, validating endpoint coverage, or helping agents onboard to the project.

Implement React components, pages, hooks, and UI features for the StudyPlanner web application. Use this skill for ANY frontend task: building new components, editing existing ones, forms, file uploads, result displays, routing, or TypeScript types. Also use when the user asks about folder structure, styling conventions, or where to put a new file. This skill enforces the agreed project structure, TypeScript conventions, and React best practices — always consult it before writing any frontend code.