Research Paper Kb

PassAudited by ClawScan on May 1, 2026.

Overview

The skill is a coherent research-paper note-taking helper, with the main thing to notice being that it creates persistent workspace memory about papers.

This appears safe for its stated purpose. Before installing, be comfortable with it sending paper titles/IDs to arXiv or Semantic Scholar and with it creating persistent PAPERS.md and MEMORY.md files in the workspace. Review those files occasionally, especially if your research topics are confidential.

Findings (2)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Paper identifiers or titles may be sent to academic metadata services, and PAPERS.md may be created or changed in the workspace.

Why it was flagged

The skill instructs the agent to make external metadata requests and modify workspace files. This is disclosed and central to the paper-KB purpose, but users should know it can create or update files.

Skill content
Also fetch the Semantic Scholar API...; Check if `PAPERS.md` exists in the workspace root. If not, create it... **Append** (never overwrite)
Recommendation

Use it for papers you are comfortable looking up through arXiv/Semantic Scholar, and review PAPERS.md changes before sharing or committing the workspace.

What this means

Saved paper summaries, threat levels, and research-overlap notes may persist and be reused in later sessions.

Why it was flagged

The skill intentionally stores and reuses persistent context across sessions. This is aligned with the stated purpose, but persistent notes can reveal research interests and influence future agent behavior.

Skill content
Updates `MEMORY.md` with a pointer so future sessions know the KB exists; ... infer from prior PAPERS.md entries and MEMORY.md
Recommendation

Periodically review PAPERS.md and MEMORY.md for accuracy, remove sensitive or outdated entries, and avoid storing confidential research context unless the workspace is trusted.