Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    cohort and segmentation explorer

    1

    This skill automates the generation of acquisition-period cohort tables and retention matrices from entity and event data.

    $2.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    cohort and segmentation explorer

    cohort and segmentation explorer

    Example session with this skill installed

    Analyze the customer_data.csv and events.log files as of 2023-12-31. Use a monthly grain, 12-month window, and compare segments by the 'signup_channel' attribute. Band by event count.

    • Read your context and instructions
    • Compiled the cohort and segmentation
    • Generated the data export

    The analysis of the provided dataset is complete. The acquisition cohorts and retention matrix have been calculated using the monthly grain for the 12-month window requested.

    Cohort and Segmentation Explorer — TechCorp, 2024-01-05

    1. Scope, sources and assumptions

    • Sources: E1 (customer_data.csv), E2 (events.log)
    • Dataset ID: TC-RET-2023
    • Grain: month | Window: 12 | Min group size: 5
    • As-of date: 2023-12-31

    2. Cohort retention matrix

    CohortSize+0+1+2+3
    2023-011,2001.00000.45000.40000.3800
    2023-029501.00000.42000.3900n/a (suppressed)

    3. Pooled retention curve

    OffsetCohortsActiveDenominatorRetention
    01212,00012,0001.0000
    1114,80011,0000.4364

    4. Segment comparison

    SegmentSizeRetention (+3)95% CIMean Events
    Organic5,0000.4200+/- 0.0212.4
    Paid4,5000.3500+/- 0.038.2

    5. Behaviour bands

    Band NameMin EventsCount% of Total
    Regular102,5000.2500
    Occasional16,0000.6000
    Dormant01,5000.1500

    6. Segment definitions and approval

    • High-Retention Organic: Entities from signup_channel 'Organic' with > 10 events.
    • At-Risk Paid: Entities from signup_channel 'Paid' with 0 events in +3 period.

    7. Exclusions, warnings and limitations

    • Excluded 45 records due to missing consent.
    • Warning: 12 entities had events dated before acquisition; these were ignored.

    8. Interpretation and items for human review

    • Observed: Organic signups show a 7% higher retention at month 3 compared to paid.
    • Analytical estimate — validation required: The dip in Feb cohort may correlate with the site outage.
    • Management decision required: Approval needed for 'At-Risk' segment definition before export.

    This analysis supports, and does not replace, professional judgment and management's decision.

    Next steps

    1. Validate the 'At-Risk' segment criteria with the product team.
    2. Review the excluded records to ensure consent flags are being captured correctly.

    customer_data.csv

    CSV · data export

    Generated

    Example file from a real run - the skill writes it into your workspace.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    This skill automates the generation of acquisition-period cohort tables and retention matrices from entity and event data. It solves the problem of manual, error-prone retention tracking by using a deterministic script to calculate counts, rates, and intervals without leaking future information into historical cohorts. The tool handles small-group suppression for privacy and ensures only consented data is processed.

    What it does

    • Cohort Analysis builds acquisition-period tables with fixed denominators to track entity activity over time.
    • Retention Curves generates pooled retention data across multiple cohorts to identify long-term trends.
    • Segment Comparison compares groups based on approved attributes, providing retention rates and 95 percent intervals.
    • Behavioral Banding categorizes entities into rule-based bands like dormant, occasional, or regular.
    • Data Validation enforces strict exclusions for records without consent or events dated before acquisition.

    How it works

    1. Source Verification registers entity and event sources, confirming dataset approval and consent flags.
    2. JSON Preparation builds a request.json file following the schema for cohort rules, grains, and attributes.
    3. Deterministic Execution runs the included Python script to calculate the retention envelope.
    4. Report Generation produces a structured Markdown report with cohort tables, curves, and behavior bands.

    Frameworks & tools

    The skill utilizes a self-contained Python execution environment (scripts/run.py) and standard JSON schema validation. It is designed for environments that support Python 3 and local file manipulation.

    Why this beats prompting it yourself

    Prompting an LLM to calculate retention rates often results in "hallucinated" percentages or flawed logic regarding future-lookahead bias. This skill offloads calculations to a deterministic script, ensuring every number is mathematically grounded and audit-ready.

    Use cases

    • Generating a monthly retention matrix for a subscription-based user base.
    • Comparing retention differences between specific marketing channels or attributes.
    • Defining behavioral segments for user engagement reporting.
    • Suppressing small-group data points to maintain internal reporting privacy.

    Known limitations

    Does not support predictive modeling, causal inference, or path-based funnel analysis. Requires pre-pseudonymized entity IDs and does not process direct identifiers like emails.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.

    Requires Python 3. No support for predictive modeling or causal inference. Requires pre-pseudonymized data. Entities below the minimum group size are suppressed.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 3 days ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables
    Listed3 days ago

    What's inside

    Frequently Asked Questions

    Unlike LLMs that often hallucinate percentages, this skill uses a deterministic Python script to calculate retention rates. This eliminates "lookahead bias" and ensures your cohort tables are mathematically accurate and audit-ready.

    The skill requires two primary inputs: an entity table (identifying when a user joined) and an event table (tracking user actions). Both must be provided in a structured format with clear consent flags and pseudonymized IDs.

    You receive a self-contained execution environment including the core Python calculation script, JSON schemas for configuration, and a documentation guide on how to register your data sources.

    Yes, the skill includes built-in small-group suppression and strictly enforces exclusions for records without consent. It is designed to work with pseudonymized IDs to ensure individual privacy is maintained in your reporting.

    The skill is optimized for environments that support Python 3 and local file manipulation. It can be integrated into most AI agent frameworks that allow for custom script execution and JSON-based request handling.