Batch Marketing Copy: a Tokeny case study
Generate diverse promotional copy for multiple channels (Weibo, Xiaohongshu, WeChat Moments) at once. 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.
- product-brief.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.
为同一产品分别写微博、小红书、朋友圈推广内容，各 3 条，适配长度和语气；输出含渠道、文案、行动引导的 CSV 和审阅文档。 交付文件：copy.csv、copy.md。 验收要点： - 共 9 条文案 - 没有虚构背书或新增功能 - 三个渠道的文案有区别 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），每份 400–800 字，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。
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
Input
This task used the files available in the workspace: product-brief.md and ENVIRONMENT.md. The product is “Songguo Task Board,” a fictional demonstration product for teams of 3–10 people. The supplied product facts are limited to task lists, assignees, due dates, CSV export, a free plan for up to 3 people, a team plan priced at RMB 29 per person per month, and the contact address hello@example.test. The brief explicitly says that the product has no mobile app, AI scheduling, or third-party certification, and forbids invented customer testimonials, sales figures, or endorsements. ENVIRONMENT.md was used only as environment context.
Processing
I first read the workspace inputs and extracted the allowed capabilities, limits, pricing, audience, and contact details. I then adapted the copy to three channels. Weibo copy uses compact problem-solution framing and a relatively high information density. Xiaohongshu copy uses a more conversational, practical, and step-based style that can support saving and sharing. WeChat Moments copy is shorter and more personal, with a team-collaboration context. I produced three entries per channel, with a separate call to action for every entry, and wrote them to copy.csv. I also wrote the review notes to copy.md, plus this Chinese report and the English report.
Actual result
The workspace now contains copy.csv, copy.md, report.md, and report.en.md. The CSV has 10 lines: one header line and nine copy records. Each channel has exactly three records. The copy uses only product facts present in the brief. It does not add a mobile app, AI scheduling, certification, customer proof, sales claims, or third-party endorsement. The channel versions differ in sentence structure, scenario framing, tone, and call-to-action wording.
Checks
I inspected the generated files to confirm that all four deliverables exist, that the CSV has the three required columns, and that the channel counts are 3/3/3. I compared every product claim against the brief and checked specifically for prohibited concepts such as mobile support, AI auto-scheduling, certification, customer testimonials, sales volume, and endorsements. I also checked that the price, user limit, and contact address match the source text. The review document records the UTF-8 CSV assumption and the channel rationale.
Limitations
No external source was used, and no content was posted or sent to any recipient. I did not verify a live landing page, current pricing, platform-specific final character limits, hashtag requirements, or advertising disclosure rules. Since the source is a fictional demonstration brief, the output does not establish real-world performance. A product owner should recheck the facts, destination URL or contact flow, and each platform’s current publishing requirements before release.
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.