Files
lanshu b639b622a7 chore: reduce repository leak surface
Stop tracking generated reports that are already ignored and replace local file:// links with relative paths.
2026-05-28 07:48:14 +08:00

1.8 KiB

Schedule Morning Global Scan

The goal is to automate a global news scan at 6:30 AM daily. Since the AI agent cannot wake up to process the data, we will implement a "headless" reporting pipeline that fetches data and generates a readable Markdown report automatically.

User Review Required

Note

AI Analysis Limitation: The automated report will contained "raw" summaries fetched from the source. It will not contain the "Deep Dive" insights (Core Value, Inspiration, Scenarios) that I usually provide, as that requires my active intelligence. Action Required: You currently need to approve the crontab command to install the schedule.

Proposed Changes

Scripts

[NEW] daily_scan.sh

  • A bash script acting as the entry point.
  • Activates the Python virtual environment.
  • Runs fetch_news.py with --source all --limit 15 --deep.
  • Pipes the output to a temporary JSON file.
  • Calls generate_basic_report.py to convert JSON to Markdown.
  • Saves the report in a date-based subdirectory reports/YYYY-MM-DD/ with a timestamped filename.

[NEW] generate_basic_report.py

  • A simple Python script.
  • Reads JSON input from stdin or file.
  • Formats it into the standard SKILL.md Markdown style (excluding the AI Deep Dive section).
  • Outputs to stdout or file.

System Configuration

  • Use crontab to schedule scripts/daily_scan.sh at 06:30 daily.

Verification Plan

Automated Execution Test

  • Run bash scripts/daily_scan.sh manually.
  • Verify a new file reports/YYYY-MM-DD/global_scan_HHMM.md is created.
  • Verify the content of the markdown file is readable and contains news from multiple sources.

Manual Verification

  • Check crontab -l after installation to confirm the job is listed.