capture-hubtoday-chrome-dev
Capture HubToday Chrome Dev OpenAI Claude
Overview
Archive the current day's HubToday AI资讯, plus yesterday's Chrome Developers Blog, OpenAI Blog, and Claude Blog updates, into the fixed daily folder under /Users/yangyilin/docs/ai-list. Use $agent-reach's web/RSS path instead of browser automation.
Quick Start
- If the environment is new or failing, run
agent-reach doctor. - Run
python3 scripts/capture_daily_sources.py. - Use
--date YYYY-MM-DDto rerun a historical archive date, or--root /tmp/...for a dry verification outside the real archive tree. - Inspect the output using
references/output-format.md.
Workflow
1. Resolve the target date and output paths
- Default the archive date to yesterday in
Asia/Shanghai. - Create
/Users/yangyilin/docs/ai-list/YYYY-MM-DD/if it does not exist.
More from jimliuxinghai/jim-skills
capture-x-ai-lists
Capture the last 1 hour of tweets from the two fixed X/Twitter lists (`https://x.com/i/lists/2013541376945013033` and `https://x.com/i/lists/1278784207641284609`) through `$agent-reach`'s Twitter/X channel and its `xreach` upstream CLI instead of the OpenClaw browser, inspect tweet images by calling the MCP `understand_image` tool through MCPorter, open only GitHub links and Article cards, use Reader first and then the OpenClaw browser for `x.com/i/article` pages when needed, and save hourly markdown records under `/Users/yangyilin/docs/ai-list/YYYY-MM-DD/HH-00.md`. Use when asked to collect these AI lists into local markdown archives instead of writing a thematic summary.
1summarize-ai-list-daily
Summarize all markdown files under yesterday's /Users/yangyilin/docs/ai-list/YYYY-MM-DD folder into daily.md. Use when asked to produce a daily AI/news summary from the ai-list archive with priority on technical depth, architecture, engineering practice, tool evaluation, and other high-signal content, including strong single-source items.
1