Goal

• Given one or more links <LINKS>, find which file(s) among the following are related to each link, and add a TODO to the most related file:
  • research/ideas/*.md
  • website/docs/blog/posts/*.md

  // TODO(gp): incorporate the link <LINK> to <TOPIC>
  

Workflow

Step 1: Collect Material

• Read the title / content of each link in <LINKS> (use WebFetch if the URL alone is not descriptive enough to judge topic)
• List all candidate files:
  • research/ideas/*.md
  • website/docs/blog/posts/*.md

Step 2: Match Links to Files

• For each link, compare its topic against the content of each candidate file (title, headers, first paragraph) and rank candidates by relevance
• Pick the single best-matching file for each link
  • If several files are close in relevance, list them and ask the user which one to use
  • If no file is a plausible match (topic not covered anywhere), do not force a match: report this to the user instead of guessing

Step 3: Add the TODO

• In the matched file, add one line:

  // TODO(gp): incorporate the link <LINK> to <TOPIC>
  

  where <TOPIC> is a short (< 15 words) description of what part of the link is relevant (e.g., a technique, a result, a quote)
• Placement:
  • If the file already has a top-of-file // TODO(gp): ... block, add the new TODO as an adjacent line in that block
  • Otherwise, add it near the section of the file the link is most related to (e.g., right above the relevant bullet point or paragraph)
• Do not change any other content of the file
• If multiple links map to the same file, add one TODO line per link

Step 4: Report

• Print, for each link, which file it was added to and the exact TODO line that was inserted
• For any link that could not be matched, report that explicitly instead of silently skipping it