decision-log

Files (3)

• README.md
• SKILL.md
• output-schema.json · references/

SKILL.md (first lines)

# Decision Log

Take signal scan output and operator preferences, run them through the Signal-to-Decision Pipeline (SDP), and produce a structured decision record. Scores opportunities across 6 dimensions, applies strategic filters, checks for cognitive bias, runs a pre-mortem, and outputs a fully logged decision with kill criteria and a next-step hint for persona-extract.

## Step 0: Load Conventions

Before doing ANYTHING, read the shared conventions file:

```
Read ${SKILLS_DIR}/_conventions.md
```

This file defines: canonical vault path, folder-to-type mapping, frontmatter contract, valid statuses, tag hierarchy, cross-reference syntax, and the PipelineEnvelope schema. All output from this skill MUST conform to those conventions. If there is a conflict between this SKILL.md and `_conventions.md`, the conventions file wins.

## Pipeline Position

```
signal-scan -> **decision-log** -> persona-extract -> offer-scope -> hunter-log
```

This skill consumes the output of `signal-scan` and feeds into `persona-extract` for deep persona research on the chosen opportunity.

## When to Use

- Choosing which opportunity to pursue from a completed signal scan
- Recording a product decision with full rationale and alternatives
- Applying structured scoring to signal-derived opportunities
- Running a pre-mortem before committing to build
- Setting kill criteria and revisit dates for a product bet

## Trigger Phrases

- "Which opportunity should I pursue?"
- "Help me decide what to build from this scan"
- "Run a decision log on [signal scan output]"
- "Log a decision for [opportunity]"
- "/decision-log [signal scan output]"

---

## Prerequisites

Before starting, the following must be available:

1. **Signal scan output** -- A completed signal-scan result containing: scan metadata, signals (all 7 types), scored opportunities with ship candidates, and a meta-signal synthesis
2. **Chosen opportunity** -- Which opportunity from the scan the operator wants to pursue (or ask them to choose after presenting the ranked list)
3. **Rationale** -- Why this opportunity over others (the operator's reasoning, beyond the score)
4. **Constraints** -- Concrete limits: "must ship in 1 day", "PDF only", "under $50", etc.
5. **Alternatives rejected** -- Which other opportunities were considered and why they lost

If only the signal scan is provided, present the ranked opportunities and guide the operator through selection before proceeding.

---

## Input

The skill expects the following input structure (typically assembled from signal-scan output and operator conversation):

```typescript
interface DecisionLogInput {
  signal_scan: SignalScanOutput       // full scan output
  chosen_opportunity: string          // title from opportunities array
  rationale: string                   // why this one
  constraints: string[]               // "must ship in 1 day", etc.
  alternatives_rejected: {
    title: string
    reason: string
  }[]
  confidence: "low" | "medium" | "high"
  revisit_by: string                  // ISO date
}
```

---

## Workflow

```
Signal Scan Output + Operator Preferences
    |
Phase 1: Frame the Decision (what exactly are we deciding?)
    |
Phase 2: Present the Options (from scan opportunities, ranked by score)
    |
Phase 3: Apply SDP Scoring (6-dimension RICE-adapted framework)
    |
Phase 4: Strategic Filters (reversibility check, two-way door bias)
    |
Phase 5: Bias Check (wanting vs. data, confirmation bias audit)
    |
Phase 6: Pre-Mortem (imagine failure -- why did it happen?)
    |
Phase 7: Make the Call (record chosen + rejected + rationale)
    |
Phase 8: Set Kill Criteria (when to abandon)
    |
Phase 9: Generate Next-Step Hint (pre-fill persona-extract inputs)
    |
Output: JSON record + Markdown summary -> vault
```

---

## Phase 1: Frame the Decision

**What exactly are we deciding?**

Do not jump to scoring. State the decision explicitly:

- What is the time horizon? ("which opportunity to prototype this week")
- What is the success metric? ("first paying customer within 2 weeks")
- What are the hard constraints? (ship time, format, budget, audience access)

**Produce a decision frame:**

```
Decision: [one-sentence statement of what we are choosing]
Time horizon: [when the bet resolves]
Success metric: [how we know it worked]

From Peleke/brunnr, MIT.

More from Peleke/brunnr