CSV Data Quality Auditor for Student Research

    1

    For students and junior analysts who receive messy CSV files, it runs a scripted audit of missing values, duplicate rows, key conflicts, and sentinel codes, and outputs a severity-ranked quality checklist.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    CSV Data Quality Auditor for Student Research

    Example session with this skill installed

    Run a quality audit on data/user_logs.csv using user_id as the id_col.

    • Read your context and instructions
    • Compiled the csv data quality
    • Generated the data export

    Audit complete for data/user_logs.csv. I ran the internal audit script and mapped the results to the standard quality template.

    CSV Data Quality Report

    1. File Overview

    • Path: data/user_logs.csv
    • Total Rows: 12,450
    • Total Columns: 5
    • Columns: user_id, timestamp, event_type, session_duration, ip_address

    2. Quality Checklist

    Issue TypeSeverityColumnCountSample Rows (Locators)
    Missing ValuesERRORsession_duration412Rows 45, 102, 330
    Duplicate RowsWARNGlobal12Rows 1202, 1203
    Type MismatchWARNtimestamp3Row 892 (invalid format)
    Sentinel ValuesINFOsession_duration85Value "-1" detected

    3. Checked & Passed

    • Primary Key Uniqueness: user_id is unique across all rows.
    • Schema Consistency: No trailing delimiters or row length mismatches detected.
    • Encoding: Verified utf-8-sig compliance.

    4. Recommended Manual Next Steps

    1. Investigate the 412 rows with missing session_duration to determine if the tracking script failed.
    2. Cross-reference the 85 instances of "-1" in session_duration with the logging documentation to see if this represents a timeout.
    3. Deduplicate the 12 redundant rows identified in the global check before importing to the DB.

    Next steps

    • Verify if the session_duration nulls should be treated as zero or discarded.
    • Check the source system for the three malformed timestamps.

    user_logs.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

    Marketplace Listing — Agensi Submission Copy

    Product Title (≤ 8 words)

    CSV Data Quality Auditor for Student Research

    One-line Pitch

    For students and junior analysts who receive messy CSV files, it runs a scripted audit of missing values, duplicate rows, key conflicts, and sentinel codes, and outputs a severity-ranked quality checklist.

    Who It Is For

    • Undergraduate and graduate students in statistics, economics, or social science who work with survey or administrative CSV exports for coursework and theses
    • Junior data analysts who need a quick go/no-go check before handing a file to pandas or SPSS

    Not intended for: production data pipelines, database-quality monitoring, or anyone who wants the data auto-fixed.

    3 Key Features (verifiable)

    1. Scripted detection, not model estimates — every number in the report (row count, missing count per column, duplicate-row count) comes from a real script run, so any figure can be re-verified by running the same command again.
    2. Six entry-error checks in one pass — missing values including N/A/null/NaN text codes, fully-duplicated rows, duplicate primary keys, text inside numeric columns, sentinel values like -999, plus empty/duplicate/whitespace column names.
    3. Read-only by design — the auditor never writes, deletes, or overwrites the input file (verified by before/after file hashing), so the original data always stays intact for review.

    One Example

    Input (a 10-row course CSV, --id-col student_id):

    student_id,name,class,pre_score,attendance_rate,homework_rate
    S001,张三,A,85,0.92,0.95
    ...
    S008,周十,B,unknown,0.91,0.90
    S010,郑二,C,-999,0.76,0.80
    

    Output summary (excerpt from the actual checklist):

    ERROR — duplicated_key: student_id has 1 duplicate (S005, rows 5–6) ERROR — duplicated_rows: 1 fully-duplicated row (rows 5–6) WARN — type_inconsistency: pre_score row 8 = "unknown" WARN — sentinel_values: pre_score row 10 = "-999" WARN — missing_values: homework_rate 2/10 rows empty Plus a "not detected" section listing every check that came back clean, and suggested manual next steps.

    Price

    Free.

    (Rationale: course-stage release. A paid tier at US$2–4 could later add multi-file batch mode and a styled HTML report, but v1.0 stays free to gather real usage feedback first.)

    Files must be under 200MB and have fewer than 500 columns. Only supports CSV format. Cannot modify, clean, or repair data.

    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

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed today

    • Free with an Agensi account

    Permissions

    Terminal / Shell
    Read Files
    Listedtoday

    What's inside

    Frequently Asked Questions

    Unlike standard LLM prompts that guess at data distributions, this skill executes a Python-based audit script on your actual file to provide mathematically precise row counts, null values, and duplicate flags.

    The skill is designed to work with any AI agent capable of executing code or running external scripts, including popular LLM frameworks that support Python environments.

    This tool is strictly an auditor; it identifies and reports errors like type mismatches and missing values but does not modify your original CSV or perform automated cleaning.

    The audit supports files up to 200MB and 500 columns, ensuring high performance while handling standard datasets and complex vendor exports.

    You receive the core audit script, the configuration ruleset for severity mapping, and the integration prompt used to generate the standardized Markdown reports.

    Yes, the auditor is compatible with both `utf-8-sig` and `gbk` encodings, making it suitable for international datasets and various legacy export formats.

    New in LLM Ops