AGENT SKILL

report-to-csv

Extract Markdown tables or bullet lists into CSV for spreadsheets.

使用场景电子表格
License 许可
MIT
所属仓库
qingjiuzys/anycode
最近更新
2026-09-12
Stars
★ 2
Forks
1

01 前言 Front Matter

mode
executable
name
report-to-csv
name_zh
报表转 CSV
version
1.1.0
approval
writes-workspace
category
research
description
Extract Markdown tables or bullet lists into CSV for spreadsheets.
permissions
{ "network": false, "read_dirs": [ "workspace" ], "write_dirs": [ "workspace" ] }
description_zh
从 Markdown 表格或列表提取数据并导出为 CSV。
channel_capabilities
[ "files", "artifacts" ]

02 资源文件 1

  • 其他run1.6 KB

03 SKILL.md 正文

report-to-csv

中文:从 Markdown 表格或列表导出 CSV。
English: Extract Markdown tables or bullet lists into CSV for spreadsheets.

Use when the user wants spreadsheet-friendly output from a report or summary.

Workflow

  1. Locate source files with Glob / FileRead (Markdown, text, or agent output).
  2. Prefer the bundled run script for the first Markdown table:
    • Skill tool → report-to-csv with args path/to/report.md [out.csv]
  3. For multiple tables or custom columns, use Bash + Edit to refine CSV manually.
  4. Validate row/column counts before delivering the path (WeChat / cron users appreciate a concrete file path).

Output

  • Default output: same basename as input with .csv extension.
  • Use UTF-8. Quote fields that contain commas.

Quality contract

  • Preserve the source column order and do not silently drop malformed rows.
  • Report source table count, exported row count, column count, and output path.
  • Empty input produces a clear error rather than an empty success artifact.

Failure recovery

  • When no Markdown table exists, show the detected headings/lists and ask which structure should become rows.
  • For inconsistent columns, write a diagnostic report and keep the source unchanged.

04 同仓库更多 Skills