- Home
- Skills
- Research & Analysis
- insightflow data intelligence
More screenshots
Works with the AI Agents you already use
insightflow data intelligence
Turn CSV and Excel datasets into clear, quality-checked analytical insights and reports.
Free
insightflow data intelligence
Example session with this skill installed
Analyze sales_q4.csv to find why margins are dipping in the Northeast region. Check for data quality issues first.
- Read your context and instructions
- Compiled the insightflow data intelligence
- Generated the data export
I have completed the initial profile and quality audit of sales_q4.csv. I identified 142 rows with missing discount codes and 12 duplicate transaction IDs that may be inflating volume metrics.
Data Quality & Analysis Report: sales_q4.csv
Dataset Overview
- File Name:
sales_q4.csv - Shape: 4,200 rows, 12 columns
- Key Columns:
transaction_date,region,category,revenue,cost,discount_pct
Data Quality Findings
| Issue | Affected Columns | Count | Recommendation |
|---|---|---|---|
| Missing Values | discount_pct | 142 | Review if NULL equals 0% or missing data. |
| Duplicates | transaction_id | 12 | Remove exact duplicates to prevent double-counting. |
| Out-of-Range | margin_pct | 5 | Flagged values > 100% (likely data entry error). |
Regional Margin Analysis (Northeast)
The Northeast region shows a mean margin of 18.4%, compared to the 24.1% average in other regions.
- Finding 1: High correlation (r=0.82) between "Northeast" and "Expedited Shipping" costs.
- Finding 2: 22% of Northeast entries have an applied
discount_pctabove the 15% threshold. - Finding 3: A significant spike in return processing fees was recorded in November.
Limitations & Assumptions
- Assumes
costincludes variable logistics but not fixed warehouse overhead. - Correlation between shipping speed and margin does not prove the speed choice caused the dip.
Next steps
- Approve the removal of 12 duplicate rows to normalize revenue.
- Decide if missing
discount_pctvalues should be filled with 0. - Provide the
logistics_fees.csvto cross-reference shipping cost anomalies.
sales_q4.csv
CSV · data export
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
Raw datasets often arrive messy, incomplete, or with hidden anomalies that skew results. This skill provides a systematic framework for profiling, auditing, and analyzing CSV and Excel files without making dangerous assumptions or silent modifications.
What it does
- Comprehensive profiling lists row/column counts, data types, and statistical summaries like mean, median, and quartiles.
- Data quality auditing flags missing values, duplicate rows, inconsistent casing, and invalid date formats for review.
- Cleaning recommendations generates a transparent plan for handling errors before any destructive changes occur.
- Pattern discovery identifies correlations, time trends, and categorical distributions based on user objectives.
- Anomaly investigation uses statistical rules to surface unusual values without assuming they are errors.
How it works
- Inspection starts by parsing the file, identifying sheets, and recording the basic shape of the data.
- Profiling generates a deep dive into every column to identify distributions and distinct values.
- Audit and cleaning presents a proposed transformation list for user approval to ensure data integrity.
- Analysis applies descriptive statistics and grouped summaries to answer specific business questions.
- Reporting delivers a structured analysis-report.md file documenting findings, limitations, and evidence.
Frameworks & tools
Works with standard tabular data formats including CSV and XLSX. Designed for environments with Python data science stacks like pandas, numpy, and matplotlib.
Why this beats prompting it yourself
Generic prompts often lead to "black box" cleaning where rows are deleted or values imputed without your knowledge. This skill enforces a strict verification workflow that requires approval for destructive changes and mandates evidence-based reporting over LLM guesswork.
Use cases
- Auditing a new customer dataset for quality issues before CRM import.
- Summarizing sales trends and categorical performance from monthly Excel exports.
- Identifying statistical outliers in financial logs for further manual investigation.
- Preparing a clean, documented dataset for downstream machine learning or BI tools.
Known limitations
Does not support file execution if the agent environment lacks a runtime. Cannot interpret instructions embedded inside dataset cells or comments.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Research & Analysis.
Requires a runtime environment for file parsing. Does not support non-tabular formats. Cannot verify causation or perform automated external data enrichment.
How to use
- 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
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 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
This skill has been used 2 times.
Trust & safety
Security scan
Security scan passed 5 days ago
- Free with an Agensi account
- Software
- Microsoft Excel
Pythonpandas
Requires an AI agent with file access and Python-based data analysis capabilities. Results depend on the tools and file formats supported by your connected agent. Always retain your original dataset before applying cleaning changes.
Permissions
File Scopes
Read access is needed to inspect user-provided CSV and Excel datasets. Write access is needed only when creating approved analysis outputs or reports. The skill does not require network access or external account connections.