HN Digest

${var} — Topic filter for stories (e.g. "rust", "LLMs", "biotech"). If empty, uses default interest areas.

Read memory/MEMORY.md for context. Read the last 2 days of memory/logs/ to avoid repeating stories.

Steps

1. Fetch top stories from the HN API:

   bashCopy

   # Get top 30 story IDs
   STORY_IDS=$(curl -s "https://hacker-news.firebaseio.com/v0/topstories.json" | jq '.[0:30][]')
   # Fetch each story's metadata
   for ID in $STORY_IDS; do
     curl -s "https://hacker-news.firebaseio.com/v0/item/${ID}.json"
   done
   

2. Filter stories by relevance. If ${var} is set, prioritize stories matching that topic. Otherwise show the top stories by points — no topic filter applied. Also include anything with 200+ points regardless of topic.

3. For the top 5-7 stories:

   • If the story has a URL, fetch it with WebFetch for more context
   • Write a one-sentence summary
   • Include the HN discussion link: https://news.ycombinator.com/item?id=ID

4. Send via ./notify (under 4000 chars):

   Copy

   *HN Digest — ${today}*
   
   1. [Title](url) (X pts, Y comments)
      Summary of why it matters.
      [Discussion](https://news.ycombinator.com/item?id=ID)
   
   2. ...
   

5. If SUPERNOTES_API_KEY is set, save the digest as a Supernotes card:

   bashCopy

   curl -s -X POST "https://api.supernotes.app/v1/cards/simple" \
     -H "Api-Key: $SUPERNOTES_API_KEY" \
     -H "Content-Type: application/json" \
     -d "$(jq -n \
       --arg name "HN Digest — ${TODAY}" \
       --arg markup "$DIGEST_MARKUP" \
       --arg color "blue" \
       '{name: $name, markup: $markup, color: $color, tags: ["hn-digest", "aeon"]}')"
   

   Where $DIGEST_MARKUP is the full digest in markdown (same content sent via notify, but with proper markdown links). If the API call fails, log a warning and continue — don't block the rest of the skill.

6. Log to memory/logs/${today}.md. If no relevant stories found, log "HN_DIGEST_OK" and end.