See it in action
One of 2 test cases the skill fixed. Same model, same prompt, with and without the skill.
PromptAfter presenting a cohort readout, executives requested deeper root-cause investigations. What template tool guides these follow-up analyses?
What was checked, without and with the skill| Checked | Without skill | With skill |
|---|
| Uses a diagnostic checklist for follow-up analyses | Missed | Passed |
|---|
What's inside
The instructions your model reads when the skill loads.
SKILL.md25 linessingle file
Cohort Analysis Framework Skill
When to Use
- Comparing performance across acquisition channels, segments, or product lines.
- Diagnosing conversion drop-offs within specific booking/vintage cohorts.
- Stress-testing forecast assumptions with historical baseline behavior.
Framework
- Cohort Definition – choose cohort key (signup month, lead source, product tier, segment).
- Metric Stack – select KPIs (coverage, win rate, ACV, NRR, payback) per cohort.
- Normalization – adjust for seasonality, deal size mix, or currency.
- Visualization – waterfall tables, heatmaps, or overlapping curves to highlight divergence.
- Narrative Layer – annotate drivers, anomalies, and recommended actions.
Templates
- Cohort definition worksheet (keys, filters, inclusion/exclusion rules).
- Standardized chart pack for leadership readouts.
- Diagnostic checklist for follow-up analyses.
Tips
- Keep cohorts mutually exclusive to avoid double-counting.
- Pair with
inspect-pipeline-levers to link cohort insights to pipeline stages.
- Rebaseline quarterly so assumptions stay current.
Track record
Benchmark
19 / 22
test cases passed, vs 18 / 22 without
Adoption
- Users
- 0
- Times used
- 0
Safety and source details
SkillSafety
✓ 3 of 3 passed✓Code scan15 of 15 checks clean
0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026
View 15 scanner checks
live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing
✓IntentSafe
LLM judge · does the skill do what it claims
✓ContentClean
LLM judge · policy & harmful-content review
Source
Importedgtmagents / gtm-agents
plugins/revenue-analytics/skills/cohort-analysis/SKILL.md
License Apache-2.0Imported 8/3/2026Version v1