How do I install the So What Sharpener skill?

Run npx skillmds@latest add daddia/so-what-sharpener in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is So What Sharpener safe to use?

This skill has not completed SkillMD's automated safety review yet. Capability flags: docs only. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with So What Sharpener?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is So What Sharpener free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published So What Sharpener?

daddia (@daddia) published this skill. Their other Agent Skills are listed on their SkillMD profile.

So What Sharpener by daddia — Agent Skill

Name: So What Sharpener
Rating: 4.2 (5 reviews)
Author: daddia

So What Sharpener

This skill should be used when the user asks "what's the so-what here," "sharpen this insight," "this is just facts, what does it mean," or provides a list of observations or data points that need to be turned into insight statements.

daddia Updated Sep 22, 2026

File contents

So-What Sharpener

When to use

Sharpen observations into insight via observation → implication → insight chain. Smallest Minto unit before full narrative.

What this skill does not do

Does not skip implication step — catches lazy insights.
Does not force insight on non-load-bearing facts — "No insight earned" section.
Does not build full pyramid — use narrative-builder for that.

Preconditions

Input	If missing
Observations or data points	Ask user to list
Decision context	Ask what decision insights must support

Provisional mode

Missing decision context: implications literal only; insights tagged [review].

Trust spine

Analytical Rigor (mandatory): Explicit three-step chain per point; merge redundant insights.
Per trust-conventions.md on tagged figures in observations.

Workflow

For each observation:

State observation unchanged.
State implication (literal).
State insight (so-what for decision).
Flag observations with no earned insight.
Group and merge overlapping insights.

Output format

Observation: [...]
  → Implication: [...]
  → Insight: [...]

No insight earned: [...]

Worked example

Input: "Mid-tier ARPU down 12% QoQ." Decision: pricing revert?

Excerpt: Implication: mid-tier spending less; Insight: increase likely priced out intended segment — reconsider pricing [review].

Quality checks before delivering

Three-step chain per point
No insight earned section when applicable
Overlaps merged

Propose profile update

When a stable convention surfaces during this run (thresholds, naming, tone, output format, or recurring corrections), propose a profile update: show the exact diff against ~/.claude/plugins/config/claude-for-strategy/consulting/CLAUDE.md (org-wide facts go to org-profile.md), ask for confirmation, and write only on yes. Only /consulting:practice-setup auto-applies a full profile write.

Outputs

Follows plugin CLAUDE.md § Outputs. Next: /consulting:narrative-builder to assemble insights.

---name: so-what-sharpenerdescription: This skill should be used when the user asks "what's the so-what here," "sharpen this insight," "this is just facts, what does it mean," or provides a list of observations or data points that need to be turned into insight statements.---# So-What Sharpener## When to useSharpen observations into insight via observation → implication → insight chain. Smallest Minto unit before full narrative.## What this skill does not do- **Does not skip implication step** — catches lazy insights.- **Does not force insight on non-load-bearing facts** — "No insight earned" section.- **Does not build full pyramid** — use `narrative-builder` for that.## Preconditions| Input | If missing ||---|---|| Observations or data points | Ask user to list || Decision context | Ask what decision insights must support |## Provisional modeMissing decision context: implications literal only; insights tagged `[review]`.## Trust spine- **Analytical Rigor (mandatory):** Explicit three-step chain per point; merge redundant insights.- Per `trust-conventions.md` on tagged figures in observations.## WorkflowFor each observation:1. State observation unchanged.2. State implication (literal).3. State insight (so-what for decision).4. Flag observations with no earned insight.5. Group and merge overlapping insights.## Output format```Observation: [...] → Implication: [...] → Insight: [...]No insight earned: [...]```## Worked example**Input:** "Mid-tier ARPU down 12% QoQ." Decision: pricing revert?**Excerpt:** Implication: mid-tier spending less; Insight: increase likely priced out intended segment — reconsider pricing `[review]`.## Quality checks before delivering- [ ] Three-step chain per point- [ ] No insight earned section when applicable- [ ] Overlaps merged## Propose profile updateWhen a stable convention surfaces during this run (thresholds, naming, tone, output format, or recurring corrections), **propose a profile update**: show the exact diff against `~/.claude/plugins/config/claude-for-strategy/consulting/CLAUDE.md` (org-wide facts go to `org-profile.md`), ask for confirmation, and write only on yes. Only `/consulting:practice-setup` auto-applies a full profile write.## OutputsFollows plugin `CLAUDE.md` § Outputs. Next: `/consulting:narrative-builder` to assemble insights.

Frequently asked questions