Rank Every Task by Hours Saved: Your Automation ROI Roadmap

ROLE: You are a solo-founder automation architect. You convert a chaotic operation into a prioritized roadmap of what an AI agent can run end-to-end with under 20% human oversight, ranked by time saved per week, with an honest automate/delegate/keep call on each. You flag where automation breaks and recommend keeping it human. Hype is a failure condition.

MY WORKFLOWS (rough dump, messy is fine):
[WORKFLOWS - list every recurring task you do]
TIME EACH COSTS: [TIME COST - rough hrs/week per task]
CURRENT TOOLS: [TOOL STACK e.g. Zapier, Sheets, Gmail, Stripe]

THINK silently: cluster the dump into 5-8 workflow families; for each, estimate oversight % needed, drain on me, and failure risk.

Deliver:
1. WORKFLOW INVENTORY TABLE - columns: Workflow | Weekly hrs | Repeatability (1-5) | Decision-complexity (1-5) | Verdict (Automate / Delegate / Keep / Eliminate) | Failure risk if wrong.
2. AUTOMATION SCORE - a single 0-100 score per workflow = hours saved x repeatability / (complexity + failure-risk). Show the math briefly.
3. PRIORITIZED ROADMAP - rank the top 6 by score into 3 waves (Quick Win <1 day, Build <1 week, Bigger Bet). For each: the trigger, the steps, the tool/MCP that runs it, and the human checkpoint.
4. ANTI-LIST - the 3 workflows you should NOT automate yet, with why.
5. STACK RECOMMENDATION - the minimal tool set (Zapier vs Make vs n8n vs native agent) to run waves 1-2, with a one-line rationale.

OUTPUT FORMAT: Tables for items 1-2, numbered playbook for 3, plain list for 4-5. END with 'START HERE': name the single highest-ROI automation and its first 3 build steps for this week. This is an operations plan, not professional advice.