LaunchKit · 2026
Back to Skills

research-paper-kb

Persistent cross-session knowledge base for research papers.

1
474 downloads
by @ilkhamfy

Setup & Installation

openclaw skills install @ilkhamfy/research-paper-kb

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install research-paper-kb

What This Skill Does

Maintains a persistent markdown file (PAPERS.md) that stores structured summaries of research papers ingested from arXiv URLs, DOIs, or titles. Each entry includes the method, claimed gap, key results, BibTeX, and a threat-level score relative to the user's own research. Entries survive across sessions because they are written to disk, not stored in context.

Unlike saving papers in chat history or a browser bookmark, entries are written to a versioned file in your workspace and remain queryable across every future session.

When to use it

  • Tracking papers that compete with your thesis work
  • Building a BibTeX library incrementally while reading literature
  • Assessing how much a newly published arXiv paper overlaps with your research
  • Reviewing all saved papers by threat level before a grant submission
  • Looking up what you previously noted about a paper without re-reading it

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User asks: Add this to my KB: https://arxiv.org/abs/2310.06825

AGENT
  1. 1Fetches abstract and metadata from arXiv for ID 2310.06825
  2. 2Queries Semantic Scholar API for citation count, TLDR, and related papers
  3. 3Extracts method, claimed gap, and key results from the abstract
  4. 4Reads MEMORY.md to infer user's research focus and assigns a threat level 1-5
  5. 5Generates a BibTeX entry with key Jiang2023mistral and appends the full structured entry to PAPERS.md
  6. 6Updates the Research Paper KB section in MEMORY.md with current paper count
OUTPUT

Confirmation summary with paper title, authors, threat level score with reason, BibTeX key, and total papers now tracked in PAPERS.md