| name | openmark-newsletter-roundup |
| description | Compose a categorical news-roundup newsletter from Ahmad's OpenMark saves. Use when Ahmad wants a scannable "what happened in AI / tools / research / discourse" recap rather than a single-thesis essay. Trigger phrases include "roundup", "news roundup", "this week in AI", "weekly news", "what dropped this week", "category recap", "/newsletter-roundup". Output is bucketed by topic with 3โ5 items per bucket, short blurbs, source-tagged for at-a-glance scanning. |
| metadata | {"type":"composition"} |
OpenMark โ Newsletter Roundup
Categorical news-roundup format. NOT thesis-driven โ this is a scannable recap of what happened, organized by topic bucket. Reader skims headers, dives into 1-2 items.
When to use vs other newsletter skills
| User intent | Skill |
|---|
| "What's happening in AI this week" | this one (roundup) |
| "Make me a newsletter on context engineering" | openmark-newsletter (analytical) |
| "Write me a thoughtful essay on X" | openmark-newsletter-essay |
| "Compare A vs B" | openmark-newsletter-comparison |
| "I want a LinkedIn-style post" | openmark-newsletter-thread |
Tool sequence
- Get time window. Default last 7 days. Ahmad can override.
- Pull recent fresh:
find_recent(days=N, n=30) โ empty query.
- Cross-source angle:
search_semantic(query="<broad area>", n=20) (whatever broad area Ahmad named, e.g. "AI tooling, models, research").
- Source-specific gems (parallel):
search_linkedin(query=..., n=10) for hot takes, search_youtube(query=..., n=10) for video drops.
- Cluster. Bucket the deduped hits into 3โ6 topical groups:
- "Models & research" โ papers, model releases
- "Tooling & open source" โ github repos, libraries, frameworks
- "Industry & business" โ funding, partnerships, strategy
- "Discourse & takes" โ LinkedIn posts, opinions
- "Tutorials & courses" โ YouTube videos, blog tutorials
- "Off-topic but interesting" โ wildcards worth a click
Don't force all 6 โ only use buckets that have โฅ2 items.
Output format โ NON-NEGOTIABLE
# {Week of YYYY-MM-DD} โ {Punchy 4-word headline}
_Quick recap of {N} items I saved {window}. {1-sentence pulse on the week.}_
## {Bucket 1 name} ({count})
1. **{Title}** โ [{domain}]({URL})
{1-sentence so-what. Why this matters.}
2. **{Title}** โ [{domain}]({URL})
{1-sentence so-what.}
(repeat 3โ5 items per bucket)
## {Bucket 2 name} ({count})
(same pattern)
## Sources cited
1. [{title}]({URL})
2. [{title}]({URL})
...
_{N} items ยท {hours} of content ยท feed pulled {window}_
Formatting rules โ strictly enforced
- One blank line between every numbered item. Otherwise the chat collapses them into one wrapped paragraph.
- Each item: title in bold, then
โ, then a short clickable [{domain}]({URL}) link (NOT the full URL pasted as text), then a blank line, then a single-sentence so-what indented or as its own line.
- Bucket headers are
## โ never ### or bold-only.
- The trailing
## Sources cited is mandatory so the chat auto-saves the draft.
Voice
Tight, declarative, no hedging. Each so-what is ONE sentence. No "this could be useful," no "I found this interesting." Examples:
| Generic | Roundup voice |
|---|
| "This is an interesting paper about RAG." | "Argues every production RAG should use cross-encoder rerank. Has numbers." |
| "A new library for agents was released." | "Drop-in replacement for LangChain's create_agent. 40% fewer LOC, same hooks." |
| "Worth checking out." | "Skip unless you're shipping agents this week." |
What NOT to do
- Don't editorialize ("the week was interesting") โ let the items speak.
- Don't pick more than 5 items per bucket. If a bucket has 10, split into two sub-buckets or drop the bottom 5.
- Don't include URLs you didn't see in a tool result.
- Don't summarize at the end. The buckets ARE the summary.
