Skills Data Science Generate ArXiv Daily Research Digest Deck

Generate ArXiv Daily Research Digest Deck

v20260803
meta-arxiv-daily-digest-deck
This meta-skill fetches the top academic papers from the arXiv API in a specified category. It generates a structured digest for each paper, renders this digest into a professional PowerPoint (PPTX) presentation (one slide per paper), and finally saves the comprehensive digest to long-term memory for future reference. Designed for daily 'morning briefings' of academic research.
Get Skill
398 downloads
Overview

{{ outputs.fetch_arxiv | truncate(10000) }} ---

      For each paper, output this exact block (separate papers with a line of `---`):

      ## <Paper Title>
      **Authors**: <first 3 author names, then et al. if more>
      **arXiv ID**: <id>
      **Core claim** (1 sentence): <claim>
      **Method summary** (2-3 sentences): <how>
      **Key numbers** (if present): <metric: value, metric: value>
      **Why it matters** (1 sentence): <relevance>
- id: render_deck
  kind: agent
  skill: pptx
  depends_on: [digest_papers]
  with:
    task: |
      Render the per-paper digest below into a PPTX deck. One slide per
      paper. Slide title = paper title. Slide body = the rest of that
      paper's digest block. Title slide first with text
      `arXiv Daily Digest — {{ inputs.user_message | xml_escape | truncate(80) }}`.

      Digest body:
      ---
      {{ outputs.digest_papers | truncate(12000) }}
      ---

      Save to: `{{ inputs.workspace_dir }}/arxiv-daily/digest.pptx`
      (overwrite OK; cron mode keeps only the latest, manual fire uses
      the same name and the user can rename after if needed). Create
      the `arxiv-daily/` subdirectory if missing.

      Reply with the absolute output path on a single line, no preamble.

      If the digest is `DIGEST_SKIPPED`, reply: `RENDER_SKIPPED`.
- id: persist
  kind: tool_call
  tool: memory_save
  tool_allowlist: [memory_save]
  depends_on: [digest_papers, render_deck]
  tool_args:
    path: "memory/arxiv-daily.md"
    mode: append
    content: |
      === arxiv-daily digest ===
      invocation: {{ inputs.user_message | xml_escape | truncate(200) }}
      deck: {{ outputs.render_deck | truncate(200) }}
      digest:
      {{ outputs.digest_papers | truncate(8000) }}

arXiv Daily Digest Deck (Meta-Skill)

A near-pure tool_call + linear DAG meta-skill: fetch arXiv via the public Atom API, write a per-paper structured digest, render it into a PPTX deck, and persist the digest into long-term memory under the arxiv-daily topic.

Trigger surface

Fire manually with the English trigger arxiv daily digest or one of the localized triggers listed in the frontmatter. Category override: include a cs.XX token anywhere in the invocation.

Fallback

If fetch_arxiv fails (network or parse), the downstream steps short-circuit by emitting _SKIPPED markers; no half-baked PPTX or memory note is written. Operator can retry by re-invoking, or manually run curl 'http://export.arxiv.org/api/query?…'.

Info
Category Data Science
Name meta-arxiv-daily-digest-deck
Version v20260803
Size 4.57KB
Updated At 2026-08-04
Language