Scheduled News Digest: a Tokeny case study
Set up scheduled tasks to auto-search industry news daily and generate summary pushes. 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.
- sources.txt
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.
实现资讯摘要脚本，读取两条官方资讯源，保存标题与 URL。使用 Tokeny 的 scheduler 工具创建临时定时任务，让它在当前时间约 2 分钟后实际触发脚本并生成 digest.md。根据当前日期设置具体分钟、小时、日和月的 Cron，不调用 trigger 代替时钟触发。读取 executions 核对任务已执行，保存真实任务 ID、计划时间、执行时间和状态到 schedule-proof.json；完成后禁用并删除这一个测试任务，不影响其他任务。另给出每日 09:00 的配置说明，但不启用长期任务、不向外部用户推送。如果调度无法运行，报告准确记录原因。 交付文件：digest.md、collect.py、schedule.md、schedule-proof.json。 验收要点： - 本次真实抓取并保存来源 - 有实际时钟触发的调度记录 - 测试任务结束后禁用并删除 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；表格的标题、表头和第一条数据保持同页，单行不要跨页拆分；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。临时网络服务请用系统分配端口（0），记录实际端口，不使用固定端口，避免并发案例互相访问。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 example uses a temporary Tokeny scheduled task and official public news sources. The report and schedule-proof.json record its trigger and cleanup results.
Results and verification
Inputs
- Workspace input:
sources.txt. - Two public official sources:
https://www.python.org/blogs/andhttps://nodejs.org/en/blog. - Run date and timezone: 2026-11-07, Asia/Shanghai.
Processing
- Wrote
collect.pyto read the first two non-empty URLs fromsources.txt, fetch them over HTTPS, and extract the page title, final URL, HTTP status, and fetch time. - Created a one-time Cron schedule,
10 14 7 11 *, and waited for the clock to fire at 14:10. The schedulertriggeraction was not used. - Read scheduler executions and confirmed that the script actually ran and produced
digest.md. - Recorded the real task ID, planned time, execution times, status, and cleanup in
schedule-proof.json. - Disabled and then deleted this test task. A final task-list check returned no tasks. No other task was touched.
Actual Results
- Execution status:
completed. - Task ID:
6443a3f9-c7cd-424e-94ad-d0ed990fd75a. - Planned time: 2026-11-07 14:10:00 +08:00.
- Started: 2026-11-07 14:10:00.021 +08:00; completed: 14:11:28.480 +08:00.
- Python.org: HTTP 200, title
Our Blogs | Python.org. - Node.js: HTTP 200, title
Node.js — Blog. - Artifact:
digest.mdwas generated with both titles, URLs, statuses, and raw records. - The test task was disabled and deleted; the final task list was empty.
Verification
- Read
digest.mdand checked both URLs, titles, and HTTP 200 responses. - Read scheduler executions and checked the execution ID, status, start time, and completion time.
- Read
schedule-proof.jsonand checked the proof fields. - Listed scheduler tasks and confirmed that the test task no longer exists.
Daily Configuration Note
For a long-running daily 09:00 schedule in Asia/Shanghai, use 0 9 * * * to run python collect.py. This configuration was documented only: no long-term task was enabled and no external recipient was contacted.
Limitations
This run records the HTML <title> and entry URL for each official blog page; it is not a full article-level content summary. Network responses, titles, and availability can change over time. On failure, the script writes a truthful failure record and exits nonzero. No temporary network service was created.
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.