decision-logger

Decision Logger

The user makes 200 micro-decisions a week. Most evaporate. The ones that matter at month 6 are usually the ones the user cannot remember making.

Decision-logger captures the non-trivial decisions, the reasoning behind them, and the date to revisit. Six months later, when the same shape of decision comes up, the log shows what the user decided last time and what they learned. This is how working-parent solopreneurs stay coherent across months of fragmented time.

When this skill triggers

Fire when the user:

• Says "logged"
• Says "let's lock this"
• Says "decided"
• Says "locked"
• Says "log this decision"
• Says "log it"
• Completes a challenge (pre-mortem) session
• Completes a hard-call session
• Has made a non-trivial decision in the current conversation (climb to paid tool, new hire, price change, channel launch, etc.)

Also fire when:

• The user proposes a decision similar to one logged in the past 90 days (surface the prior decision for context)
• risk-gauge flags a move (log the decision either way, override or hold)
• free-first-checker documents a climb

What counts as non-trivial

The skill logs decisions that meet at least one of these criteria:

• Recurring financial impact above $50/mo
• One-time financial impact above $500
• Time commitment over 10 hours
• Reputation or relationship impact (announcing publicly, ending a contract, naming a price)
• Strategic direction (offer change, channel addition, pricing structure)
• Risk above the user's stated cap

Skip:

• Tactical micro-decisions ("should I email this prospect today")
• Tool clicks ("what time should I schedule this post")
• Decisions made within a logged decision's scope (the logged decision already covered it)

When in doubt, ask the user: "Is this worth logging?"

The log structure

Each decision gets its own file at decisions/YYYY-MM-DD-<slug>.md.

---
date: YYYY-MM-DD
slug: <kebab-case-slug>
type: <hire | spend | offer | pricing | channel | quit | climb | other>
status: <active | revisited | reversed | superseded>
revisit_date: YYYY-MM-DD
related_decisions: <slug1, slug2>
related_skills: <challenge, hard-call, risk-gauge, free-first-checker>
---

# <Decision Title>

## Decision

<One paragraph. What was decided, in plain language. Include specific numbers and dates.>

## Context

<Why this came up. What state the business was in. What the trigger was.>

## Alternatives considered

- **Option A: <name>** - <one-sentence summary of the option> - <why not>
- **Option B: <name>** - <one-sentence summary> - <why not>
- **Option C: <chosen>** - <one-sentence summary> - <why yes>

## Expected outcome

<What the user expects to happen if this decision is right. Specific signals at 30, 60, 90 days.>

## Kill signal

<What would tell the user this decision was wrong. Specific. Numeric where possible.>

## Reasoning

<The actual reasoning behind the call. Quote the user's own words from the session if available. Include what they were avoiding naming if hard-call surfaced it.>

## What I am avoiding

<Optional. If the decision is being made knowing it conflicts with a stated value or cap, name the conflict. This is the honest-mirror line. Future-self will thank you.>

## Revisit checklist

When this decision is revisited at `revisit_date`, check:

- [ ] Did the expected outcome land?
- [ ] Did the kill signal fire? If so, what was done about it?
- [ ] Did the decision conflict with what I was avoiding? Was that worth it?
- [ ] What did this decision teach me about how I make decisions?

## Tags

<#tag1 #tag2 #tag3 - for cross-referencing similar decisions>