Skill UI

Tycana keeps task context across Claude conversations, capturing new work items, planning the day, tracking patterns, and offering energy-aware recommendations that deepen with each sync through MCP.

This skill automates the extraction of actionable items, commitments, and follow-ups from meeting notes (e.g., Granola, Grain). It creates structured tasks in Todoist, updates task status, and drafts follow-up emails. Ideal for running daily or post-meeting cron jobs to ensure all commitments are tracked and executed.

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.

This skill generates a comprehensive daily digest summarizing all critical pending items into a single summary. It aggregates stalled scheduling threads, unanswered emails, pending introductions/follow-ups, open Todoist tasks, and upcoming calendar events from multiple sources. It is essential for running a daily summary cron job or when the user requires a quick status check on their workload and commitments.

This tool generates comprehensive briefing documents for scheduled meetings. It integrates data from calendar systems, email history (up to 90 days and historical), professional networking sites (LinkedIn), and past interaction records (Granola). It provides context on attendees, relationship history, key discussion points, and critical action items to ensure users are fully prepared for every call, regardless of whether it's a first introduction or a follow-up.

This skill monitors Todoist for tasks that are due or overdue and require follow-up (email, ping, etc.). It automatically drafts highly contextual emails by combining task descriptions, retrieved meeting transcripts (from Granola/Grain), and prior email history. Drafts are created directly in Gmail and the user is notified via WhatsApp, saving significant time on manual follow-ups and ensuring context-aware communication.

Internal Feishu/Lark notification utility that other skills call to push event cards or run interactive dialogs via webhook or bridge, ensuring safe fire-and-forget delivery whenever checkpoints, experiments, reviews, or pipelines change state.

Automates the entire process of converting LaTeX source files into a professional, submission-ready PDF. This tool handles multi-pass compilation (using latexmk), diagnoses common LaTeX errors (such as missing packages, undefined references, and BibTeX syntax errors), and performs post-compilation checks, including page count verification according to specific academic guidelines (ICLR, IEEE, etc.). Ideal for researchers submitting papers.

This system functions as a persistent, disk-based memory for complex workflows. It structures and tracks multi-step projects and research by utilizing three key Markdown files: task_plan.md (main plan), findings.md (research discoveries), and progress.md (session log). It ensures that critical context, progress state, and key findings are never lost to the limited context window, making it ideal for long-term, structured work.

Guides users through saving their Discord bot token, checking access policies, and locking down channel reach by managing pairing codes, allowlists, and policy transitions so the Claude assistant can reliably respond via Discord.

Deep research and scientific writing assistant that produces LaTeX/BibTeX documents with verified citations for papers, proposals, reports, and reviews, ensuring well-formatted, citation-backed outputs for academic or clinical contexts.

A structured system for managing complex, multi-step tasks and research projects by using persistent Markdown files as external memory. It enforces systematic planning, progress tracking, and critical finding logging across sessions, ensuring that core knowledge is never lost in the ephemeral context window. Ideal for engineering workflows, deep research, and multi-stage development.