Skill UI

Base Refly ecosystem skill that routes user intents to domain-specific skills, runs workflows via the CLI, and automates multi-step tasks or pipelines by delegating execution to the Refly backend while providing CLI-specific triggers and commands.

Guides teams through intentional brainstorming before implementation, surfacing intent, constraints, and design options, then formalizing a spec so every creative change starts with validated direction.

This skill transforms flawed writing — whether pasted text or uploaded documents — into publication-ready prose. It eliminates grammatical, spelling, punctuation, clarity, and tone issues while strictly preserving the author’s original intent and unique voice. Use it for professional document refinement, academic paper polishing, or general text review, ensuring the corrected version maintains your authentic style.

This skill automatically drafts professional email replies for multiple corporate accounts. It handles various communication intents, including intro acceptances, scheduling follow-ups, acknowledgments, and positive replies. It operates strictly in draft-only mode, ensuring drafts are ready for user review but never sent automatically. It is ideal for automating responses triggered by new emails, webhooks, or action item follow-ups.

Manage Feishu tasks and tasklists through intent-based actions: create, query, update, complete, reopen, and organize assignments with role-aware members and ISO 8601 deadlines.

This advanced tool acts as a specialized prompt engineer, guiding users to create highly optimized, production-ready prompts for various AI models (LLMs, Midjourney, etc.). It enforces strict writing rules, guides detailed intent extraction (Task, Context, Constraints), and provides model-specific best practices (e.g., Claude vs. GPT-5.x) to maximize AI output quality and maintain consistency. Essential for professional content creators, developers, and researchers needing precise AI interaction.

The single command entry point for founders' strategic queries. It automatically analyzes keywords and intent (e.g., 'burn rate' or 'pipeline') to route the question to the most specialized C-level advisor (CFO, CRO, CMO, etc.), or convene a multi-role boardroom if multiple functions are involved. It eliminates the cognitive burden of needing to remember multiple complex slash commands.

This skill facilitates structured, bidirectional reading and knowledge synthesis for academic texts, theories, and articles. It forces users to first capture the source material's original intent, before critically mapping its relationships, conflicts, and absorptions against the CrossFrame framework. It ensures structured output with clear source boundaries, making it ideal for deep learning and scholarly work.

This skill functions as an optional agent companion for MDPR, a deterministic Markdown-to-presentation runtime. It helps review presentation workflows by providing semantic hints, visual evidence reports, and high-level layout intentions. Crucially, it keeps agent output evidence-based while letting MDPR maintain control over final slide geometry, rendering details, and layout coordinates.

A comprehensive guide and checklist for preparing submissions for top-tier academic conferences, such as POPL. This resource helps researchers adhere to strict formatting rules (e.g., 25-page limits, acmsmall), maintain full double-blind anonymity, prevent unintentional author identity leaks, and manage the complex workflow and deadlines for academic publishing.