- Home
- Skills
- Business & Operations
- Loona Sheet Rescue
Loona Sheet Rescue
Send the ugliest spreadsheet you own — get back a clean, fixed, formula-working file you can actually trust, with the original never overwritten.
Free
Works with the AI Agents you already use
Loona Sheet Rescue
Example session with this skill installed
I have a messy CRM export called 'june_leads.csv'. It has duplicate rows, weird date formats, and the 'Total Value' column isn't summing correctly. Can you rescue it using the Loona Method?
- Read your context and instructions
- Compiled the loona sheet rescue
- Generated the data export
I have initiated the Loona Sheet Rescue Method for june_leads.csv. I have preserved the original and completed the intake and preservation steps.
Diagnosis Report — june_leads.csv
Preserved original: june_leads.csv (untouched)
Working copy: june_leads-WORKING.csv
Sheets and row counts: leads_data — 154 rows
Findings:
- High Severity: 12 exact duplicate rows found in leads_data.
- Medium Severity: 'Lead Date' column contains three different formats (MM/DD/YYYY, DD-MM-YY, and text).
- Medium Severity: 'Total Value' column contains 8 values stored as text, causing the SUM formula to skip them.
- Flag-for-owner: Row 42 has a date '05/06/24' which is ambiguous (May 6 or June 5).
- Flag-for-owner: 3 near-duplicates found (similar names/emails but different phone numbers).
Will be fixed automatically: Exact duplicates will be removed, text-to-number conversion for unambiguous amounts, and date standardization.
Needs your decision: Please confirm the date format for Row 42 and whether to merge the 3 near-duplicates.
Verdict: Rescuable as scoped.
Status: Awaiting owner decisions on flagged items.
Rescued free by Loona Sheet Rescue. For month-end books that stay clean, see Loona Month-End Close; for invoice lists that chase themselves, see Loona ListingPro.
Next steps
- Tell me if '05/06/24' in Row 42 is May 6 or June 5.
- Review the 3 near-duplicates (Leads #12, #88, #92) and tell me if they should be deleted or kept.
- Once you confirm, I will proceed to the Clean and Repair stages.
june_leads.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
Messy spreadsheets with broken formulas, inconsistent dates, and untrustworthy totals create operational debt for small businesses. This skill implements the Loona Sheet Rescue Method to audit, clean, and repair Excel or CSV files without risking the original data. It follows a strict safety protocol to ensure every change is logged and every formula is verified through independent recomputation.
What it does
- Original preservation creates an identical working copy to ensure the source file remains untouched and read-only.
- Data auditing produces a comprehensive Diagnosis Report identifying duplicates, formatting errors, and broken formulas before any changes occur.
- Deep cleaning removes exact duplicates and standardizes inconsistent date, phone, and currency formats.
- Formula repair fixes
#REF!,#VALUE!, and#N/Aerrors and rebuilds up to 10 core calculation formulas. - Change logging records every single value change with old and new values, ensuring 100% auditability.
- Verification testing reconciles row counts and performs independent recomputations to prove the rescued data is accurate.
How it works
- Intake: Provide the messy file and define what each column represents and which numbers must not change.
- Diagnosis: Review the generated Diagnosis Report to approve auto-fixes or decide on flagged ambiguous data.
- Execution: The agent performs cleaning and repairs following the Loona Method (Preserve, Diagnose, Clean, Repair, Verify).
- Delivery: Receive a new
-CLEANfile alongside a full Change Log, the final Diagnosis Report, and a Tidy-Up Checklist.
Frameworks & tools
Works with .xlsx Excel files, CSVs, and Google Sheets exports.
Why this beats prompting it yourself
Standard prompting often leads to silent data loss or "hallucinated" values when an AI tries to guess missing data. This skill enforces strict Hard Rules that prevent silent changes and mandate independent verification of every repaired formula.
Use cases
- Cleaning up CRM exports with mixed date formats and duplicate customer entries.
- Repairing invoice lists where amounts are stored as text and totals disagree with column data.
- Auditing inherited inventory sheets to identify broken lookups and range errors.
- Standardizing contact lists for email marketing or sales outreach.
Known limitations
Limited to one workbook per rescue. Supports up to 10 formula repairs or additions per run. Does not support password-protected or corrupted files, macro authorship, or live database connections.
One workbook per rescue. Max 10 formula repairs. No password-protected or corrupted files. Does not author macros or handle live shared workbooks. Flags but does not interpret ambiguous data.
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 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 3 days ago
- Free with an Agensi account