- Home
- Skills
- Productivity
- Image Table to CSV
Works with the AI Agents you already use
Image Table to CSV
Visual table transcription from images to CSV, with merge-cell handling, append mode for multi-page documents and validation.
Free
Image Table to CSV
Example session with this skill installed
Here is the table from page 3 of the scan (attachment: page-0003.png). I need it as CSV.
- Read your context and instructions
- Compiled the image table to
- Generated the data export
Read the image and wrote data/processed/page-0003.csv.
| Check | Result |
|---|---|
| Columns | region, site, beds, occupancy, source |
| Rows | 3 |
| Merged cells | North repeated on all three rows |
| Empty rows | none |
region,site,beds,occupancy,source
North,Alpha,120,87%,page-0003
North,Beta,64,91%,page-0003
North,Gamma,30,78%,page-0003
Next steps: if there are more pages of the same table, send them and I will append to the same file.
page-0003.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
Transcribes a table that exists only in an image — a photo, a screenshot, a scanned page — into a clean CSV, by reading it visually instead of running OCR.
OCR breaks on the tables that matter most: merged cells get lost, accents turn into symbols, 80 comes out as $o and 100 as 190. This skill makes the agent read the image with semantic understanding: it sees that a cell spans four rows because it is vertically centred, and it knows a percentage cannot exceed 100.
What the buyer gets:
- Column naming rules that keep CSVs from different pages concatenable.
- Merge-cell and group-heading handling, spelled out with a worked example.
- Append mode with a carry-over state, so a 12-page scanned document becomes one continuous CSV instead of twelve fragments.
- A validation pass (row count, empty rows, sequential columns,
sourcecolumn) before the file is reported as done. - Context-budget guidance: 3 to 5 images per turn, with a clean stop instead of a truncated table.
Compatibility: works with any agent that can see images. No scripts, no dependencies, no network access.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Productivity.
Requires the agent to see the image; a text-only model cannot use this skill.
Handwriting and heavily degraded scans are transcribed best-effort.
Very wide tables (more than ~15 columns) may need to be split by the user.
It does not read PDFs directly: render the page to PNG first, or use the PDF to Markdown pack.
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
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 to use
No account required to browse
Trust & safety
Security scan
Security scan passed 8 days ago
- Free with an Agensi account
An agent with image input, and a writable folder for the CSV output.
Permissions
File Scopes
Reads image files the user points at and writes CSV files to the output folder. The output may contain personal data taken from the source document; the skill instructs the agent not to copy it into the conversation and not to move the file to shared folders, and nothing in the package sends data over the network. No network access and no external binaries.