Ai Productivity Stack

Design a personal AI-enhanced workflow that saves time without creating tech debt.

Overview

AI Productivity Stack is a workflow design workshop that helps users build a personal AI-enhanced productivity system. It guides users through auditing their current tasks, identifying where AI adds genuine value, and constructing a stack for email, writing, research, scheduling, and task management — with "human-in-the-loop" checkpoints to prevent over-automation.

This skill focuses on personal productivity, not automating away jobs or responsibilities that require human judgment.

When to Use

Use this skill when the user asks to:

• Use AI for productivity
• Build an AI workflow
• Automate tasks with AI
• Find AI tools for getting things done
• Work faster with AI

Trigger phrases: "AI for productivity", "Build an AI workflow", "Automate with AI", "AI tools for getting things done", "Work faster with AI"

Workflow

Step 1 — Greet and Assess

Acknowledge the user's productivity goals. Ask:

• What does a typical workday or week look like?
• What tools do they currently use?
• What are their biggest productivity pain points? (email overload, writer's block, research time, meeting prep, task prioritization)
• How comfortable are they with trying new tools?

Step 2 — Task Audit

Guide the user through listing their regular tasks. For each task, assess:

• Frequency: How often does it occur?
• Time consumed: How long does it take?
• Cognitive load: Is it draining or mechanical?
• AI fit: Could AI help? (high/medium/low/none)

Categorize tasks into:

• High AI fit: Drafting, summarizing, brainstorming, formatting, transcribing
• Medium AI fit: Research, scheduling, note organization, first-pass editing
• Low/none AI fit: Final decision-making, sensitive communication, creative direction, relationship building

Step 3 — Design the AI Touchpoints

For high and medium-fit tasks, design specific AI touchpoints:

Email:

• AI drafts responses to routine emails
• Human reviews, personalizes, and sends
• Never AI-send without human review for important communications

Writing:

• AI generates outlines and first drafts
• Human provides voice, examples, and final polish
• AI helps with editing and clarity, not authorship

Research:

• AI summarizes long articles and reports
• Human verifies key facts and reads primary sources for critical decisions
• AI generates questions to guide deeper reading

Scheduling and Task Management:

• AI suggests time blocks based on priorities
• Human approves and adjusts
• AI generates meeting agendas; human owns the outcomes

Step 4 — Build the Stack

Recommend a lightweight, integrated stack based on the user's tools and comfort level:

• Minimal stack: One AI chatbot + existing tools (lowest friction)
• Balanced stack: AI chatbot + one specialized tool (e.g., meeting transcriber, research assistant)
• Advanced stack: Multiple AI tools with clear role separation

Emphasize: start with the minimal stack. Add tools only when a clear gap exists.

Step 5 — Human-in-the-Loop Checkpoints

Establish rules for when the human must be involved:

• Never fully automate: Sensitive emails, performance reviews, conflict resolution, any communication with emotional weight
• Always review: Anything sent externally, factual claims, data analysis
• Human owns: Final decisions, creative direction, relationship management

Create a personal "AI boundary list" — tasks the user decides AI will never touch.

Step 6 — Summarize and Exit

Recap the user's personalized productivity audit and recommended stack. Provide:

• A staged implementation plan (start with highest-impact, lowest-risk change)
• A reminder to review the stack after 2-4 weeks of use
• Suggest related skills: Prompt Library Builder for reusable productivity prompts, AI Decision Framework for prioritization

Safety & Compliance

• Focuses on personal productivity, not automating away jobs or responsibilities
• Does not encourage AI use where human judgment is critical (e.g., final editing of important communications)
• Does not recommend tools that violate platform terms of service
• Emphasizes human-in-the-loop for all external-facing work
• This is a descriptive prompt-flow skill with zero code execution, zero network calls, and zero credential requirements

Acceptance Criteria

1. User describes their work/tasks; output includes a task-by-task AI fit assessment
2. A recommended stack is provided at an appropriate complexity level for the user
3. Human-in-the-loop checkpoints are explicitly defined
4. A staged implementation plan is provided
5. Does not recommend fully automating sensitive or relationship-critical tasks

Examples

Example 1: Knowledge Worker

User says: "I spend too much time on email and meeting notes. How can AI help?"

Skill guides: Audit typical week. Identify high-fit tasks (email drafting, meeting summarization). Design touchpoints: AI drafts, human reviews. Recommend minimal stack: AI chatbot for drafting + existing note app. Set boundary: human reviews all external emails. Provide staged plan: week 1 — email drafting; week 2 — meeting notes; week 3 — review and adjust.

Example 2: Freelancer Overwhelmed by Admin

User says: "I'm a freelancer drowning in admin tasks. Can AI save me?"

Skill guides: List admin tasks (invoicing, client communication, scheduling, proposal writing). Assess AI fit for each. Design touchpoints: AI drafts proposals from templates, AI suggests schedule blocks, AI generates invoice reminders. Emphasize: client relationships stay human. Recommend balanced stack. Create boundary list.