Batch Format Conversion: a Tokeny case study
Batch convert documents between Markdown, Word, PDF, HTML, and other formats. Follow the recorded task, execution screenshots, actual files and verification notes.
By the Tokeny team · Windows · Tokeny 1.7.8 · gpt-6-sol
Prepare the inputs
Create a separate workspace in Tokeny, download and unpack the sample inputs, then select a model that supports tools. The sample uses synthetic data or the public sources named in the article.
- guide.md
- notes.md
Send the task to Tokeny
The original task was executed in Chinese. Copying these instructions reproduces the same task; the results below include an English report.
把两份 Markdown 分别转为 Word 和 HTML，保留标题、列表和表格，输出转换清单与失败记录。 交付文件：guide.docx、guide.html、notes.docx、notes.html、conversion.csv。 验收要点： - 两份输入各有两个输出 - 待确认内容未丢失 - HTML 与 Word 可打开 这是使用合成演示输入或指定公开资料的真实操作案例。请先读取工作空间输入，再实际生成文件并自检。所有写入、脚本与临时服务限于当前工作空间；临时服务只监听 127.0.0.1，执行后关闭；不触达外部收件人、不发布、不付款、不更改系统设置，不调用子智能体。需要公开资料时可读取给定 URL。只启用完成任务必需的工具。另写 report.md（中文）和 report.en.md（英文），简洁、完整，包含输入、处理步骤、真实结果、检查方式和局限；失败或未验证部分如实记录，不能宣称成功。最终回复用中文列出产物、关键结果与待核实项。 排版检查：Word 主标题仅出现一次，标题为黑色且与后续正文或表格保持同页；Excel 表头和长文字不能被窄列截断，合理设置列宽、换行、行高；HTML 在 390px 视口不能整页横向溢出。报告没有固定字数要求，不要为字数反复改写。
The recorded run used an isolated demonstration workspace. Review the commands and file paths before approving operations in your own workspace.
What actually ran
The task was submitted through the Tokeny input box. An online model read the inputs and executed tools. This run covers the local sample task and its actual files. Results and checks are recorded below.
Results and verification
Inputs
guide.md: a guide with a title, subheading, unordered list, and table.notes.md: meeting notes with a title, ordered list, and bold pending-confirmation text.
Processing
- Read both Markdown files as UTF-8.
- Parsed headings, unordered/ordered lists, tables, and plain/bold text.
- Wrote one DOCX and one responsive HTML file per input.
- Wrote
conversion.csvwith one record per input-output pair. - Performed structural and content checks, including a narrow viewport check for HTML.
Actual Results
All four requested output files were generated. conversion.csv contains four successful conversion records. The table in guide and the pending text “待确认：日期尚未确定。” in notes were retained.
Checks
- DOCX: re-read with
python-docxand checked headings, lists, tables, and pending text. - HTML: checked expected tags, tables, and pending text; inspected layout in a local browser at a 390px viewport.
- CSV: checked the header, four records, and output file existence.
Limitations
The inputs did not include code blocks, images, links, nested lists, or complex tables, so those Markdown features were not specifically tested. Final Word font rendering may vary with installed fonts and Office versions.
Download and inspect the outputs
The archive contains the files created during this run. Use these checks when adapting the example to your own data.
- Reconcile the reported results against the downloadable sample inputs.
- Open the generated files and check the contents relevant to your task.
- Review the limitations and any unverified steps in the report.