What problem does it solve?

Every AI conversation starts from zero — no memory of what you captured yesterday, no awareness of what's overdue, no sense of your work patterns. Tycana gives Claude persistent memory about your work so every conversation picks up where the last one left off.

Who uses this workflow?

Individual developers and knowledge workers who use Claude Code daily and want task management integrated into their AI workflow rather than a separate app.

What's included

A SKILL.md for the Tycana productivity backend, which connects via MCP to provide:

• Task capture from conversation — mention something you need to do and it's captured with effort, energy, and project inferred from context
• Day planning — prioritized plans based on what's due, overdue, and your actual completion patterns
• Progress reviews — weekly summaries with slip detection and pattern awareness
• Energy-aware recommendations — "what should I work on next?" factors in deadlines, priorities, and your current energy level
• Computed intelligence — velocity tracking, effort calibration, and stale work detection that improve over time

Example

User: "Plan my day — I've got about 6 hours"
Claude: pulls tasks from Tycana, factors in overdue items, suggests prioritized plan

User: "Remind me to update the deployment docs before Friday"  
Claude: captured instantly with context, no app switching

User: "Review my week"
Claude: summarizes completions, carry-overs, and patterns worth noting

Links

• MCP endpoint: https://app.tycana.com/mcp (OAuth 2.1)
• Getting started: https://www.tycana.com/getting-started/
• Plugin repo: https://github.com/tycana/tycana-claude-plugin