Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    descriptive statistics and distribution analyzer

    1

    Your data analysis workflow is only as good as its transparency.

    $2.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    descriptive statistics and distribution analyzer

    descriptive statistics and distribution analyzer

    Example session with this skill installed

    Analyze the provided Employee_Pulse_2024 dataset (approved by HR_Ops). Variables: Salary (ratio), Dept (nominal). I need summary stats and a comparison by Dept using 'Sales' as the reference.

    • Read your context and instructions
    • Compiled the descriptive statistics

    The execution for dataset Employee_Pulse_2024 is complete. The calculation script processed 150 records after excluding 2 duplicates.

    Descriptive Statistics and Distribution Analysis — Employee_Pulse_2024, 2024-05-20

    1. Dataset, approval and scope

    • Dataset ID: Employee_Pulse_2024
    • Source: E1 (Provided CSV)
    • Approved by: HR_Ops
    • Record Flow: 152 received, 2 excluded (duplicates), 150 analyzed.
    • Scope: All requested variables (Salary, Dept) were within allowed access.

    2. Numeric variables

    Variablen validMeanMedianSDIQRMADTrimmed meanSkewness G1Excess kurtosisOutliers (low/high)
    Salary15072450.0068000.0015200.5012000.008500.0071000.000.850.120 / 4

    Note: Skewness and kurtosis are adjusted sample estimators. Quantiles use inclusive linear interpolation.

    3. Categorical variables

    CategoryCountProportion
    Engineering600.40
    Sales500.33
    Marketing300.20
    HR100.07

    4. Subgroup comparisons

    Grouping: Dept

    LevelnMeanMedianSDDiff (Mean)Cohen d
    Sales (Ref)5065000.0064000.005000.000.000.00
    Engineering6082000.0081000.007000.00+17000.002.81
    Marketing3068000.0067500.005500.00+3000.000.57

    Differences are descriptive; no test was run and no cause is implied.

    5. Data quality, exclusions and suppressed cells

    • Missing values: Salary (0), Dept (0).
    • Invalid values: 0 detected.
    • Suppressed cells: None.

    6. Interpretation notes and items for human review

    Analytical estimate — validation required: The Salary distribution is moderately right-skewed (G1=0.85); the median (68,000) may be a more representative center than the mean.

    Management decision required: Four high-side outliers detected in Salary; review is required before using these figures for budget planning.

    This analysis supports, and does not replace, professional judgment and the data owner's decision.

    Next steps

    1. Review the 4 high-side outliers in the Salary variable to ensure data entry accuracy.
    2. Confirm if the 'Engineering' salary lead aligns with current market benchmarks.
    3. Obtain HR_Ops approval before sharing the subgroup comparison table.

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

    About this skill

    Your data analysis workflow is only as good as its transparency. Relying on opaque LLM-generated math leads to "hallucinated" averages and unreliable distributions. This skill enforces a deterministic, script-driven approach to summary statistics.

    What it does

    • Computes central tendency including mean, median, and trimmed mean with exact decimal precision.
    • Analyzes dispersion and shape providing variance, standard deviation, IQR, MAD, skewness, and kurtosis.
    • Generates frequency tables for categorical and ordinal data with automatic median category detection for ordered sets.
    • Performs subgroup comparisons calculating differences between groups and reference levels, including Cohen d and eta squared.
    • Enforces data privacy through small-cell suppression and strict consent/access field checks.
    • Audits data quality by flagging missing, invalid, or out-of-range values per variable.

    How it works

    1. Input validation: Provide a dataset and define variable scales (nominal, ordinal, interval, or ratio).
    2. Request generation: The agent builds a structured request.json ensuring all source materials are traceable.
    3. Deterministic execution: A Python script performs every calculation, ensuring zero "LLM math" errors.
    4. Result reporting: The agent interprets the script's output envelope, applying required markers for human review and management decisions.

    Frameworks & tools

    This skill utilizes a specialized Python calculation engine (scripts/run.py) and JSON schemas for strict input/output validation.

    Why this beats prompting it yourself

    Generic prompts often struggle with exact statistical formulas, especially robust measures like MAD or trimmed means. This skill removes the risk of LLM calculation errors by offloading math to a deterministic script while enforcing enterprise-grade privacy controls like small-cell suppression.

    Use cases

    • Analyze numeric distributions to identify skewness and Tukey-defined outliers.
    • Compare department or regional performance through descriptive subgroup tables.
    • Generate auditable frequency reports for survey data with categorical variables.
    • Screen datasets for missingness and invalid values before higher-level modeling.

    Known limitations

    Does not perform hypothesis testing (p-values), correlations, or predictive modeling. Limited to descriptive summaries of provided, approved datasets.

    No significance tests or p-values. No causal modeling or correlation analysis. Requires pre-approved datasets with defined measurement scales. No automated missing data imputation.

    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 any payment.

    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 2 days ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access
    Environment Variables
    Listed2 days ago

    What's inside

    Frequently Asked Questions

    This skill prevents 'LLM math' errors by using a deterministic Python engine for all calculations, ensuring your means, medians, and distribution metrics are mathematically exact rather than probabilistic guesses.

    The skill is compatible with any AI agent or platform that supports Python execution and JSON-based tool calls, such as OpenAI Assistants or LangChain-based frameworks.

    You receive the core Python calculation engine, the JSON schemas for input/output validation, and the configuration files required to integrate the skill into your agent's workflow.

    To protect individual privacy, the skill automatically applies small-cell suppression, meaning it will withhold or mask results where the sample size is too small to maintain anonymity.

    The skill is strictly for descriptive statistics; it does not perform inferential testing (like t-tests or p-values), correlations, or any form of predictive machine learning.

    Yes, when you provide a dataset, you can define variable scales—nominal, ordinal, interval, or ratio—to ensure the engine applies the correct statistical formulas for each data type.