文献助手

v1.0.0

literature retrieval and single-pdf analysis for academic research. use when the user asks to find papers by topic, author, keyword, or latest papers in the...

0· 119·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for lyt0302/literature-helper.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "文献助手" (lyt0302/literature-helper) from ClawHub.
Skill page: https://clawhub.ai/lyt0302/literature-helper
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install literature-helper

ClawHub CLI

Package manager switcher

npx clawhub@latest install literature-helper
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name and description (literature retrieval and single-PDF analysis) match the SKILL.md. However, the instructions assume the agent can perform web searches and fetch/confirm full-text PDFs and can parse uploaded PDFs; those capabilities are not declared in the skill metadata (no required tools or env). This is a functional gap rather than a security mismatch, but the user should confirm the agent has appropriate search/browsing and file-processing tools.
Instruction Scope
SKILL.md stays on task: it instructs how to build bilingual search strings, return structured search results, label availability, and to extract/analyze content from a single uploaded PDF. It does not direct the agent to read unrelated files, environment variables, or to transmit data to external endpoints beyond normal searching/fetching. It explicitly discourages fabrication.
Install Mechanism
No install spec and no code files — instruction-only skill (lowest install risk). Nothing is downloaded or written to disk by the skill itself.
Credentials
No required environment variables, credentials, or config paths are requested — proportional to the described functionality.
Persistence & Privilege
always is false and the skill does not request elevated or persistent privileges or modifications to other skills. Default autonomous invocation is allowed (normal for skills) but not combined with other concerning factors.
Assessment
This skill appears coherent and low-risk: it needs no secrets or installs. Before enabling, confirm the agent environment actually provides web-search/browsing and PDF parsing (so the skill can find papers and read uploaded PDFs). Be cautious about uploading PDFs that contain sensitive or unpublished data, and always manually verify any paper links, availability labels, and extracted facts (LLMs can still hallucinate despite the 'do not invent' instruction). If you need searches to come from a specific source (e.g., Google Scholar, Semantic Scholar, arXiv, institution subscriptions), verify or configure which tool or API the agent will use.

Like a lobster shell, security has layers — review code before you run it.

latestvk976zvqc8qm61v0gjbhbh10k358429sf
119downloads
0stars
1versions
Updated 3w ago
v1.0.0
MIT-0

Literature Helper

Literature retrieval

When the user asks for literature on a topic, author, keyword, research field, or latest papers:

  1. Identify:

    • topic
    • optional author
    • optional time range
    • optional language preference
  2. If the user asks for "latest literature" but gives no time range, default to the last 5 years.

  3. Generate:

    • Chinese keywords
    • English keywords
    • 2 to 4 useful search strings
  4. Search for relevant papers and return a structured result list.

  5. For each result, include:

    • title
    • authors
    • year
    • source or journal
    • one-sentence summary
    • link
    • availability label:
      • open access
      • repository pdf
      • abstract only
      • availability unconfirmed
  6. Never claim full-text availability unless it is actually confirmed.

Single PDF analysis

When the user uploads one PDF:

  1. Extract:

    • title
    • keywords
    • research question
    • methods
    • data or materials
    • main findings
    • conclusion
    • limitations
  2. Then give a clear Chinese summary unless the user asks for another language.

  3. If the user asks questions about the PDF:

    • answer primarily from the PDF
    • if the answer is not found in the PDF, say clearly that it was not found in the document
    • do not invent details

Output style

For literature retrieval, use:

Search terms

  • Chinese:
  • English:
  • Suggested search strings:

Results

  1. Title
    • Authors:
    • Year:
    • Source:
    • Summary:
    • Link:
    • Availability:

For PDF analysis, use:

Basic information

Keywords

Research question

Methods

Data or materials

Main findings

Conclusion

Limitations

Answers to user questions

Reliability rules

  • Be explicit about uncertainty.
  • Never fabricate paper links, authors, journals, or availability.
  • Distinguish clearly between:
    • open access direct pdf
    • repository version
    • abstract page only
    • unconfirmed access

Comments

Loading comments...