# Manage Lark Sheets with Reliable CLI Workflows

Complex spreadsheet work requires precise commands and careful validation. This skill guides Claude, Codex, and Claude Code through reliable Lark Sheets operations.

## Install

```bash
npx skillstore add larksuite/lark-sheets
```

## Metadata

- Status: approved
- Slug: larksuite-lark-sheets
- Version: 3.0.0
- Author version: 3.0.0
- Skillstore revision: r2
- Version status: valid
- Tree hash: 35bc6295d23077a15123268d79be7efbbca948f67c361c79873d6895c1fbb9bd
- Author: larksuite
- GitHub username: larksuite
- License: MIT
- Repository: https://github.com/larksuite/cli/tree/main/skills/lark-sheets/
- Ref: ebdfe608f5de2b66ff37ab4af12af8ac4f5e8006
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: network, external\_commands, filesystem
- Quality score: 85
- Quality tier: gold
- Public page: https://skillstore.pages.dev/skills/larksuite-lark-sheets
- Manifest: https://skillstore.pages.dev/api/skills/larksuite-lark-sheets/manifest

## Capabilities

- Create, import, export, inspect, and reorganize Lark workbooks and sheets.
- Read and write cell values, formulas, styles, comments, images, and typed table data.
- Insert, delete, resize, hide, freeze, merge, move, copy, fill, clear, and sort ranges.
- Create and maintain charts, pivot tables, filters, filter views, conditional formats, and sparklines.
- Translate Excel formulas into supported Lark Sheets formulas and verify returned results.
- Combine related write operations into ordered batch updates with dry-run and confirmation guidance.

## Use Cases

- Analyze operational data: Inspect ranges, preserve data types, calculate summaries, and build charts or pivot tables for recurring reports.
- Maintain shared trackers: Add rows, update formulas, apply consistent formatting, and verify changes without replacing unrelated content.
- Build financial models: Create typed workbooks, translate formulas, add sensitivity views, and validate calculated outputs.

## Prompt Templates

### Inspect a workbook

```
Inspect this Lark Sheets URL: {{url}}. List its sheets, identify {{sheet}}, and summarize the structure without changing data.
```

### Update a data range

```
Read {{sheet}} range {{range}}, apply {{changes}}, preserve nearby formatting and data types, then verify every changed cell.
```

### Create an analysis view

```
Analyze {{source_range}} by {{group_field}} and {{metric}}. Create a pivot table and suitable chart, then verify their source ranges.
```

### Migrate a financial model

```
Rebuild {{model_name}} in Lark Sheets from {{source}}. Preserve typed inputs, translate formulas, add sensitivity analysis, and validate representative calculations.
```

## Limitations

- The skill requires an installed and authenticated lark-cli plus its declared lark-shared sibling guidance.
- It does not search Lark Drive by file name; use lark-drive to locate unknown spreadsheets.
- Write and delete operations can change shared data, so users must review targets and confirmations.
- Some Excel formulas require translation because Lark Sheets does not support every Excel function or array behavior.

## Best Practices

- Inspect workbook and sheet identifiers before reading or writing a range.
- Use dry-run for destructive operations and verify affected ranges after every write.
- Preserve numeric and date types when data must support sorting, formulas, charts, or pivots.

## Anti Patterns

- Do not guess sheet names, identifiers, ranges, or target positions.
- Do not overwrite source columns or shared data without checking the complete affected range.
- Do not store numbers, percentages, or dates as formatted text when typed values are required.

## Security Audit

- Audited at: 2026-07-23T17:01:40.922\+00:00
- Summary: All 271 static findings were false positives caused by Markdown syntax, placeholder URLs, documentation links, JSON examples, and dense Chinese text. Two subprocess examples invoke the required lark-cli with fixed argument arrays and no shell. No prompt injection, data exfiltration, hidden payload, or unsafe runtime behavior was found.

## Stats

- Views: 150
- Downloads: 23
- Favorites: 1
- Popularity score: 0