metrics-review
Skill1 starsChanged 7 months agoMITView on GitHub
Review and analyze product metrics with trend analysis and actionable insights. Use when running a weekly, monthly, or quarterly metrics review, investigating a sudden spike or drop, comparing performance against targets, or turning raw numbers into a narrative. Trigger phrases: "metrics review", "review our numbers", "what happened to our metrics", "weekly review", "monthly metrics", "KPI review", "why did X drop", "analyze these metrics", "dashboard review", "performance review numbers"
3 other projects use a version of this.
Adopted by
Most recently taken up by
- anthropics/knowledge-work-pluginstheir own version, first seen Oct 13, 2026
- robisson/build-like-amazon-agent-skillstheir own version, first seen Oct 5, 2026
- agentdynarq/founder-skillstheir own version, first seen Jul 15, 2026
Other versions3 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 3
- anthropics/knowledge-work-plugins★ 36ksame name
- robisson/build-like-amazon-agent-skills★ 16same name
- agentdynarq/founder-skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository164 other files it gives its agents.
- instrument-data-to-allotropeSkill
- literature-reviewSkill
- nextflow-developmentSkill
- scientific-problem-selectionSkill
- scvi-toolsSkill
- single-cell-rna-qcSkill
- discover-brandSkill
- enforce-voiceSkill
- generate-guidelinesSkill
- csat-reportSkill
- customer-escalationSkill
- customer-researchSkill
- draft-responseSkill
- kb-articleSkill
- sentiment-analysisSkill
- ticket-triageSkill
- analyzeSkill
- build-dashboardSkill
- create-vizSkill
- data-context-extractorSkill
- data-modelingSkill
- data-pipelineSkill
- etl-designSkill
- explore-dataSkill
- sql-queriesSkill
- statistical-analysisSkill
- validate-dataSkill
- write-querySkill
- accessibility-reviewSkill
- component-librarySkill
- design-critiqueSkill
- design-handoffSkill
- design-systemSkill
- frontend-designSkill
- research-synthesisSkill
- user-researchSkill
- ux-copySkill
- wireframingSkill
- Agent DevelopmentSkill
- CLAUDE.md ManagementSkill
- MCP Server BuilderSkill
- Plugin DevelopmentSkill
- Prompt EngineeringSkill
- Skill CreatorSkill
- curriculum-designSkill
- lesson-planningSkill
- quiz-generatorSkill
- study-guideSkill
- tutoringSkill
- api-designSkill
- architectureSkill
- ci-cd-pipelineSkill
- code-reviewSkill
- database-migrationSkill
- debugSkill
- dependency-auditSkill
- deploy-checklistSkill
- documentationSkill
- error-handlingSkill
- git-workflowSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 3 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 18
Before looking at numbers, clarify:
- Distinctive: line 57
Is it real?: Check for data pipeline issues, tracking changes, or definition drift.
- Distinctive: line 60
Is it persistent or one-time?: Check if the trend is continuing or already reverting.
- Distinctive: line 62
Use the "five whys" to find root cause.
- Distinctive: line 77
If targets do not exist, recommend them based on historical performance and growth rate.
- Distinctive: line 143
Be honest about data quality: If a number looks suspicious, say so.
“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Product Metrics Review
Review product metrics systematically: identify trends, investigate anomalies, compare against targets, and produce actionable insights for stakeholders.
Workflow
Step 1: Establish Context
Before looking at numbers, clarify:
- Time period: What window are we reviewing? (week, month, quarter)
- Comparison baseline: Prior period, same period last year, or target/plan?
- Metric set: Which KPIs matter most? (If unknown, start with the standard set below)
- Known events: Any launches, incidents, holidays, or changes during the period?
Standard product metric set if none specified:
| Category | Metrics |
| Acquisition | New signups, activation rate, traffic by channel |
| Engagement | DAU/WAU/MAU, session frequency, feature adoption |
| Retention | D1/D7/D30 retention, churn rate, cohort curves |
| Revenue | MRR/ARR, ARPU, expansion/contraction, conversion rate |
| Quality | Error rate, load time, support ticket volume, NPS/CSAT |
Step 2: Scan for Signals
Review each metric against its baseline:
| Metric | Current | Prior | Delta | vs Target | Signal |
|--------|---------|-------|-------|-----------|--------|
| DAU | 12,400 | 11,800| +5.1% | Above | Normal |
| Churn | 4.2% | 3.1% | +1.1pp| Above | Alert |Flag metrics as:
- Alert: Moved more than 2x normal variance or crossed a threshold
- Watch: Moved more than expected but not yet critical
- Normal: Within expected range
- Positive: Improved beyond target
Step 3: Investigate Anomalies
For each Alert or Watch signal, diagnose:
- When did it start? Narrow the timeframe. Daily or hourly granularity.
- Is it real? Check for data pipeline issues, tracking changes, or definition drift.
- Is it broad or narrow? Segment by platform, geography, cohort, plan tier.
- What else changed? Correlate with deployments, campaigns, external events.
- Is it persistent or one-time? Check if the trend is continuing or already reverting.
Use the "five whys" to find root cause. Do not stop at the first explanation.
Step 4: Compare Against Targets
For each metric with a target:
### [Metric Name]
- **Target**: [Value]
- **Actual**: [Value]
- **Gap**: [+/- Value and %]
- **Trajectory**: [On track / At risk / Off track]
- **Forecast**: [If current trend continues, where do we land?]If targets do not exist, recommend them based on historical performance and growth rate.
Step 5: Produce the Review
## Metrics Review: [Period]
Prepared: [Date]
### TL;DR
[2-3 sentences: overall health, biggest win, biggest concern]
### Scorecard
[Table with all metrics, status indicators]
### Highlights
1. [Positive trend with quantification and likely cause]
2. [Positive trend with quantification and likely cause]
### Concerns
1. [Negative trend with quantification, investigation results, recommended action]
2. [Negative trend with quantification, investigation results, recommended action]
### Deep Dive: [Most Important Anomaly]
[Root cause analysis with supporting data]
### Actions
| Action | Owner | Priority | Due |
|--------|-------|----------|-----|
| [Action 1] | [Name] | P1 | [Date] |
| [Action 2] | [Name] | P2 | [Date] |
### What to Watch Next Period
- [Metric or trend to monitor]
- [Metric or trend to monitor]Cadence-Specific Guidance
Weekly reviews: Focus on short-term signals and operational health. Keep it brief. Emphasize what changed and what to do about it this week.
Monthly reviews: Include trend lines and cohort analysis. Compare to plan. Identify strategic themes, not just tactical issues.
Quarterly reviews: Step back to strategic health. Are we moving the right metrics? Should targets be revised? Include competitive and market context.
Examples
Example 1: Investigating a Drop
User: "Signups dropped 20% this week." Response: Segment by channel to isolate the source. Check paid vs. organic. Look for tracking changes or site issues. Compare day-by-day to find the exact inflection. Check if the funnel is broken (traffic down vs. conversion down). Cross-reference with any deploys or campaign changes.
Example 2: Routine Weekly Review
User: "Here are our weekly numbers: [data]. How are we doing?" Response: Build the scorecard, flag alerts and watches, provide 2-3 highlights and 1-2 concerns with recommended actions. Keep the narrative concise for a team meeting format.
Example 3: Target Setting
User: "We need to set Q3 targets for our key metrics." Response: Analyze trailing 6-month trend lines. Calculate growth rates. Propose targets at current trajectory and stretch targets at 1.5x growth rate. Note which metrics have the most variance and should have wider ranges.
Best Practices
- Lead with the story, not the spreadsheet. Executives want "retention improved because X" not a wall of numbers.
- Always compare. A number without context is meaningless. Compare to prior period, target, or cohort.
- Separate signal from noise. Weekly metrics are noisy. Use rolling averages or statistical significance tests before sounding alarms.
- Segment before concluding. A flat top-line metric can hide a growing segment and a shrinking one. Always check at least one level deeper.
- Quantify the impact. "Churn is up" is less useful than "the churn increase represents $45K ARR at risk."
- Track leading indicators. Lagging metrics (revenue, churn) tell you what happened. Leading metrics (activation, engagement) tell you what will happen.
- Be honest about data quality. If a number looks suspicious, say so. Do not present unreliable data with false confidence.
- Close the loop. Reference actions from the previous review. Were they completed? Did they have the expected effect?
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.

