metric-storyteller
Skill1 starsChanged 7 months agoMITView on GitHub
Turn raw metric snapshots into narrative explanations for stakeholders. Use when preparing updates, explaining metric movements, or making data-driven arguments.
In this repository26 other files it gives its agents.
- AGENTS.mdAGENTS.md
- app-store-analyzerSkill
- brainstormingSkill
- competitor-trackerSkill
- continual-learningSkill
- decision-loggerSkill
- exec-communicatorSkill
- experiment-designerSkill
- experiment-writeupSkill
- growth-loop-mapperSkill
- jtbd-extractorSkill
- knowledge-updaterSkill
- launch-postSkill
- launch-readinessSkill
- meeting-to-actionsSkill
- one-pagerSkill
- pptx-creatorSkill
- prd-writerSkill
- prioritization-scorerSkill
- retro-runnerSkill
- stakeholder-updateSkill
- strategy-connectorSkill
- user-interview-synthSkill
- what-ifSkill
- working-backwardsSkill
- writing-clearlySkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 2 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 16
Before stakeholder updates, board prep, or exec reviews
- Distinctive: line 17
When a metric moved unexpectedly and you need to explain it
- Very specific: line 56
Check memory/decisions/ for related product changes
- Very specific: line 57
Check memory/feedback/ for related user sentiment
- Very specific: line 79
If this analysis reveals new insights, update knowledge/products/*/metrics.md with the latest snapshot.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Metric Storyteller
Transform raw numbers into clear narratives that stakeholders actually understand. Bridges the gap between "retention dropped 2%" and "here's what happened, why it matters, and what we're doing about it."
When to Use
- "Explain these metrics to my manager"
- "Why did [metric] change?"
- "Help me tell the story of this data"
- "Prepare a metrics update"
- Before stakeholder updates, board prep, or exec reviews
- When a metric moved unexpectedly and you need to explain it
Process
1. Ingest the Data
Accept metrics in any format:
- A pasted table or spreadsheet
- Reference to
knowledge/products/*/metrics.md - A screenshot or description ("DAU dropped 8% this week")
For each metric, capture:
- Current value
- Previous period value (WoW, MoM, QoQ)
- Target (if set)
- Baseline (what "normal" looks like)
2. Classify the Movement
For each metric, classify:
| Signal | Definition | Narrative tone |
| On track | Moving toward target, within expected range | Confident, brief |
| Outperforming | Exceeding target or growing faster than expected | Celebratory, attribute cause |
| Underperforming | Below target but within tolerance | Watchful, explain drivers |
| Alert | Significant deviation, requires action | Urgent, action-oriented |
| Flat | No meaningful change | Contextual — is flat good or bad? |
3. Build the Narrative
For each metric worth discussing:
The headline — one sentence that captures the "so what"
- Bad: "Retention is 42%"
- Good: "Day-7 retention improved to 42%, crossing our 40% target for the first time since launch"
The context — why did this happen?
- Reference recent launches, experiments, or external events
- Check
memory/decisions/for related product changes - Check
memory/feedback/for related user sentiment - Note seasonality or external factors (holidays, competitor launches, OS updates)
The implication — what does this mean going forward?
- On current trajectory, where do we land by end of quarter?
- Does this change our priorities or plan?
- What's the confidence level in this trend continuing?
The action — what are we doing about it (if needed)?
- Already in progress: "We shipped X last week, expecting impact in Y days"
- Planned: "We're running experiment Z to address this"
- Needs decision: "We need to decide between A and B"
4. Tailor to Audience
Adjust depth based on who's reading:
- Exec/board — headline + implication only, 2-3 sentences per metric
- Manager/stakeholder — headline + context + action, short paragraph
- Team — full detail including methodology and caveats
5. Update Knowledge Base
If this analysis reveals new insights, update knowledge/products/*/metrics.md with the latest snapshot.
Output
## Metrics Narrative: [Product/Feature]
**Period:** [date range]
**Audience:** [exec / manager / team]
### Headline
[One sentence summary of overall health]
### Key Metrics
**[Metric 1]: [value] ([direction] [percent]% [period])**
[Status badge: On Track / Outperforming / Underperforming / Alert]
[2-4 sentence narrative: what happened, why, what it means, what we're doing]
**[Metric 2]: [value] ([direction] [percent]% [period])**
[Status badge]
[Narrative]
### Trends to Watch
- [Emerging pattern that isn't urgent yet but could become one]
### Actions in Flight
| Action | Expected Impact | Timeline |
|--------|----------------|----------|
| [action] | [metric + direction] | [when] |Consumer/Mobile Metrics to Watch
Common metrics for consumer/mobile PMs (use as a checklist):
- Acquisition: installs, organic vs. paid, cost per install
- Activation: onboarding completion, time to first value
- Engagement: DAU/MAU ratio, sessions per user, time in app
- Retention: D1, D7, D30 retention curves
- Monetization: ARPU, conversion to paid, LTV
- Virality: referral rate, k-factor, share rate
Principles
- Lead with "so what" — the number is not the story
- Attribute causation carefully — correlation is easy, causation is hard
- Always pair a problem with an action — never leave stakeholders with just bad news
- One metric, one paragraph — if it takes longer, the audience is wrong
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

