Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Retention Calculator · dicklesworthstone | DecimalAI

dicklesworthstone/retention-calculator

NewSkillScoreMaintained

Configure and manage - Calculate retention calculator operations. Auto-activating skill for Data Analytics. Triggers on: retention calculator, retention calculator Part of the Data Analytics skill category. Use when working with retention calculator functionality. Trigger with phrases like "retention calculator", "retention calculator", "retention".

by dicklesworthstone

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Data Analyst pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/dicklesworthstone-retention-calculator/SKILL.md

decimalai skills pull dicklesworthstone-retention-calculator --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/dicklesworthstone-retention-calculator/SKILL.md https://app.decimal.ai/s/dicklesworthstone-retention-calculator/SKILL.md

Raw file: /s/dicklesworthstone-retention-calculator/SKILL.md ↗ · machine-readable summary at /s/dicklesworthstone-retention-calculator.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 23d ago

No — -67% lift+16% tokens0% turns64% of whole cases pass with skill (14/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-10	✗→✓	▲ Improved	79%	0%
case-04	✓→✗	▼ Worse	11%	0%
case-05	✓→✗	▼ Worse	29%	0%
case-08	✓→✗	▼ Worse	42%	0%
case-21	✓→✗	▼ Worse	10%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/10/2026

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

Imported

dicklesworthstone / pi_agent_rust

tests/ext_conformance/artifacts/plugins-community/skills/12-data-analytics/retention-calculator/SKILL.md

View on GitHub ↗

License MITImported 8/10/2026Version v1

Skill contents

SKILL.md

Retention Calculator

Overview

This skill provides automated assistance for retention calculator tasks within the Data Analytics domain.

When to Use

This skill activates automatically when you:

Mention "retention calculator" in your request
Ask about retention calculator patterns or best practices
Need help with data analytics skills covering sql queries, data visualization, statistical analysis, and business intelligence.

Instructions

Provides step-by-step guidance for retention calculator
Follows industry best practices and patterns
Generates production-ready code and configurations
Validates outputs against common standards

Examples

Example: Basic Usage Request: "Help me with retention calculator" Result: Provides step-by-step guidance and generates appropriate configurations

Prerequisites

Relevant development environment configured
Access to necessary tools and services
Basic understanding of data analytics concepts

Output

Generated configurations and code
Best practice recommendations
Validation results

Error Handling

| Error | Cause | Solution | |-------|-------|----------| | Configuration invalid | Missing required fields | Check documentation for required parameters | | Tool not found | Dependency not installed | Install required tools per prerequisites | | Permission denied | Insufficient access | Verify credentials and permissions |

Resources

Official documentation for related tools
Best practices guides
Community examples and tutorials

Related Skills

Part of the Data Analytics skill category. Tags: sql, analytics, visualization, statistics, bi

Illustrative Want to see what dicklesworthstone/retention-calculator would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/30/2026, 07:57 UTC · 22 cases

Lift

-67%

17% vs 83%

Agent turns

1.0 vs 1.0

Tokens

+16%

4,174.27 vs 3,594.95

fail → pass · 1pass → pass · 13fail → fail · 3pass → fail · 5

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 Write a PostgreSQL SQL query to calculate N-day retention (specifically Day 0, Day 1, Day 7, and Day 30) for users in an e-commerce platform using the `users` table (`user_id`, `created_at`) and `user_actions` table (`user_id`, `action_timestamp`). Many developers mistakenly join `user_actions` on `created_at` directly without establishing user cohort start dates or fail to handle users with no activity on later days. Ensure all users from the cohort denominator are preserved.	fail→fail	28,597	31,895	+12%	1	1	0%	5,078	5,293	+4%	0	0	—
case-02 Construct a Snowflake SQL query that generates a cohort retention matrix with cohort creation week in rows and retention periods (Week 0 through Week 4) in columns for standard product usage logs in `app_events`. Analysts often rely on dynamic PIVOT clauses that fail when new weeks are introduced or write simple GROUP BY queries that output long vertical rows instead of a horizontal grid.	pass→pass	19,448	21,330	+10%	1	1	0%	2,880	4,100	+42%	0	0	—
case-03 Build a SQL query to calculate 30-day retention for mobile app user cohorts created in the past 60 days from table `user_signups` and `event_logs`. A common mistake in retention analytics is including recent signups (e.g., users who joined 5 days ago) in the denominator for Day 30 retention, which artificially inflates calculated churn.	pass→pass	18,121	21,578	+19%	1	1	0%	2,640	3,839	+45%	0	0	—
case-04 Write a SQL query in BigQuery to compute 7-day rolling retention (unbounded retention) for a SaaS platform using `signups` and `activity_logs`. A common error is using exact day equality (`DATE_DIFF(activity_date, signup_date, DAY) = 7`), which ignores users who returned on Day 8 or later.	pass→fail	16,607	16,130	-3%	1	1	0%	2,343	2,597	+11%	0	0	—
case-05 Optimize a retention SQL query targeting a multi-billion row event log table `raw_clickstream` in PostgreSQL. Querying raw logs directly via self-joins causes memory overflows and query timeouts because users log hundreds of actions per day.	pass→fail	26,148	28,999	+11%	1	1	0%	4,041	5,196	+29%	0	0	—
case-06 Provide a SQL query pattern to calculate daily cohort retention for a global app stored in `global_user_events` where `event_timestamp` is stored in UTC but business reporting requires US Eastern Time (`America/New_York`). Converting dates after truncating causes boundary events at 11 PM ET to be assigned to the wrong day.	pass→pass	18,463	18,104	-2%	1	1	0%	2,830	3,128	+11%	0	0	—
case-07 Write a SQL query for PostgreSQL calculating the product stickiness ratio (DAU / MAU) over a 30-day window for table `active_sessions`. Beginners often average daily active user counts divided by total monthly signups instead of dividing single-day DAU by unique monthly active users over the rolling 30-day period.	pass→pass	18,674	18,371	-2%	1	1	0%	2,667	3,031	+14%	0	0	—
case-08 Construct a SQL query that assigns users to cohort start dates using `user_events` table when there is no dedicated `users` or `signups` table. Beginners often use `GROUP BY user_id` in a subquery or mix non-aggregated columns, leading to syntax errors or incorrect cohort dates when calculating retention.	pass→fail	16,389	18,463	+13%	1	1	0%	2,205	3,127	+42%	0	0	—
case-09 Create a BigQuery SQL query to calculate bracketed retention where activity within Days 1-7 is Week 1, Days 8-14 is Week 2, and Days 15-21 is Week 3. A common error is using integer division without handling day zero or using overlapping `BETWEEN` date boundaries that double-count users across week boundaries.	fail→fail	20,921	22,683	+8%	1	1	0%	3,598	4,239	+18%	0	0	—
case-10 Write a data validation SQL query to test an existing retention calculation pipeline output table `cohort_retention_summary` (`cohort_date`, `period_offset`, `retained_users`, `cohort_size`). The pipeline frequently produces corrupted reports where Day 0 retention is less than 100%.	fail→pass	16,307	16,934	+4%	1	1	0%	2,284	4,087	+79%	0	0	—
case-11 Write a SQL query to calculate monthly cohort retention for a SaaS recurring subscription model using `subscriptions` (`user_id`, `start_date`, `end_date`, `status`). Beginners attempt to count discrete login events from event logs rather than testing active subscription date ranges across billing cycles.	pass→pass	16,301	22,868	+40%	1	1	0%	3,388	3,957	+17%	0	0	—
case-12 Create a SQL query that calculates the cumulative cohort churn rate and period-over-period incremental churn rate from a retention matrix table `retention_metrics` containing `cohort_month`, `month_number`, and `retention_rate`.	pass→pass	12,588	15,521	+23%	1	1	0%	2,796	2,652	-5%	0	0	—
case-13 Construct a SQL query to measure user resurrection for an online community platform using `user_daily_activity`. A resurrected user is defined as an active user in month M who was completely inactive in month M-1 but had activity prior to M-1. Analysts often misclassify new signups as resurrected users.	pass→pass	16,758	19,622	+17%	1	1	0%	3,482	4,566	+31%	0	0	—
case-14 Write a SQL query in Snowflake that formats retention cohort data from `app_usage` specifically for consumption by BI dashboard tools like Tableau as a standardized long-format dataset (`cohort_group`, `period_name`, `cohort_size`, `retained_count`, `retention_pct`).	pass→pass	20,442	20,937	+2%	1	1	0%	3,021	3,629	+20%	0	0	—
case-15 Write a SQL query in PostgreSQL to calculate 7-day retention specifically for core interaction events (where `event_name = 'completed_checkout'`) after initial signup in `users` (`user_id`, `signup_date`). Beginners often count any pageview event in the numerator, inflating retention metrics beyond actual feature engagement.	pass→pass	15,812	12,747	-19%	1	1	0%	2,972	2,837	-5%	0	0	—
case-16 Write a SQL query to calculate Return Retention (percentage of cohort users active on Day N or any day after Day N) versus Classic N-Day Retention (active exactly on Day N) for mobile app `user_sessions`.	pass→pass	22,817	29,149	+28%	1	1	0%	3,715	6,784	+83%	0	0	—
case-17 Write a SQL query to compute 1-day, 7-day, and 30-day cohort retention broken down by signup platform (`ios`, `android`, `web`) from `user_profiles` and `activity_logs`. Analysts often forget to include platform in the grouping set, aggregating all platforms together and masking platform-specific drop-offs.	pass→pass	11,710	17,001	+45%	1	1	0%	2,549	2,916	+14%	0	0	—
case-18 Write a SQL query in BigQuery to calculate retention across 14-day bi-weekly periods for subscription services in `user_activity`.	fail→fail	22,425	22,437	+0%	1	1	0%	3,866	4,045	+5%	0	0	—
case-19 Write a SQL query that generates cohort retention metrics ensuring cohorts with zero active users in subsequent periods still appear with 0 retained users and 0.0 retention percentage rather than missing rows.	pass→pass	20,721	21,415	+3%	1	1	0%	3,105	3,804	+23%	0	0	—
case-20 Build an XGBoost machine learning model pipeline in Python using `scikit-learn` and `xgboost` to predict individual customer Lifetime Value (LTV) and churn probability over the next 90 days based on historical transaction features.	pass→pass	40,910	34,020	-17%	1	1	0%	7,887	6,455	-18%	0	0	—
case-21 Write an Apache Airflow DAG in Python that orchestrates a daily ETL pipeline to extract raw clickstream events from AWS S3, clean the json records using PySpark, and load them into a PostgreSQL data warehouse.	pass→fail	24,172	29,740	+23%	1	1	0%	5,018	5,509	+10%	0	0	—
case-22 Design a real-time event streaming pipeline architecture using Apache Kafka and Redis pub/sub to track user click events in high-throughput web applications with sub-millisecond ingestion latency.	pass→fail	37,022	35,557	-4%	1	1	0%	6,724	6,043	-10%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of -67 percentage points is the difference between those two pass rates over the 22 comparable cases. 5 cases got worse with the skill loaded, and they are included in that figure.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 83.3%
Pass rate with the skill: 16.7%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 5
Skill version measured: v1
Run completed: 8/30/2026, 07:57 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-10

Prompt

Write a data validation SQL query to test an existing retention calculation pipeline output table `cohort_retention_summary` (`cohort_date`, `period_offset`, `retained_users`, `cohort_size`). The pipeline frequently produces corrupted reports where Day 0 retention is less than 100%.

Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.

Show both answers

SkillScore blends 4 signalsno signals yet

BenchmarkRun a verified benchmark with positive lift

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+18% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

No signals yet — run a benchmark or collect evaluated traces to earn a score.

Verified on gemini-3.6-flash · re-tested 22 days ago

Other measured skills in the registry, with their headline benchmark lift.