Source: https://github.com/aipoch/medical-research-skills

Research Article Weekly

This skill generates a weekly report of academic literature for a given keyword. It searches PubMed for articles published in the last 7 days, classifies them into generic research categories (Fundamental, Applied, Methodology, Review, Other), and produces a summarized report. This tool is domain-agnostic and adapts to any research field indexed in PubMed.

When to Use

• Use this skill when you need generates a weekly academic literature report based on keywords using pubmed. use when the user wants to track recent research progress on a specific topic, automatically retrieving, classifying, and summarizing relevant papers from the last 7 days in a reproducible workflow.
• Use this skill when a evidence insight task needs a packaged method instead of ad-hoc freeform output.
• Use this skill when the user expects a concrete deliverable, validation step, or file-based result.
• Use this skill when

  scripts/pubmed_search.py

  is the most direct path to complete the request.
• Use this skill when you need the

  research-article-weekly

  package behavior rather than a generic answer.

Key Features

• Scope-focused workflow aligned to: Generates a weekly academic literature report based on keywords using PubMed. Use when the user wants to track recent research progress on a specific topic, automatically retrieving, classifying, and summarizing relevant papers from the last 7 days.
• Packaged executable path(s):

  scripts/pubmed_search.py

  .
• Structured execution path designed to keep outputs consistent and reviewable.

Dependencies

• Python

  :

  3.10+

  . Repository baseline for current packaged skills.

• Third-party packages

  :

  not explicitly version-pinned in this skill package

  . Add pinned versions if this skill needs stricter environment control.

Example Usage

cd "20260316/scientific-skills/Evidence Insight/research-article-weekly"
python -m py_compile scripts/pubmed_search.py
python scripts/pubmed_search.py --help

Example run plan:

1. Confirm the user input, output path, and any required config values.
2. Edit the in-file

   CONFIG

   block or documented parameters if the script uses fixed settings.
3. Run

   python scripts/pubmed_search.py

   with the validated inputs.
4. Review the generated output and return the final artifact with any assumptions called out.

Implementation Details

See

## Workflow

above for related details.

• Execution model: validate the request, choose the packaged workflow, and produce a bounded deliverable.
• Input controls: confirm the source files, scope limits, output format, and acceptance criteria before running any script.
• Primary implementation surface:

  scripts/pubmed_search.py

  .
• Parameters to clarify first: input path, output path, scope filters, thresholds, and any domain-specific constraints.
• Output discipline: keep results reproducible, identify assumptions explicitly, and avoid undocumented side effects.

Inputs

• keywords

  : The search term(s) for the literature (e.g., "lung cancer", "CRISPR", "machine learning", "climate change").

Workflow

1. Search & Generate Draft Report: The skill executes the bundled Python script to search PubMed and generate a draft Markdown report using rule-based classification. This provides an immediate, usable output even without LLM processing.

   python scripts/pubmed_search.py --keywords "{keywords}" --days 7 --limit 20 --format markdown
   

2. Refine Report (Optional - AI Enhanced): If an AI environment is available, the Agent can take the raw JSON output (by running with

   --format json

   ) or the draft Markdown report and refine it using the advanced logic below for better summarization and topic extraction.

   Classification Logic (for AI Refinement): For each retrieved article (Title + Abstract), classify it using the following logic:

   System Prompt:

   Act as a versatile research analyst. Your task is to categorize the research article based on its title and abstract in relation to the user's keyword: "{keywords}".

   Research Types:

   1. Fundamental Research: Theoretical studies, mechanisms, basic science, discovery, or foundational work.
   2. Applied Research: Practical applications, clinical trials, engineering implementations, case studies, or field deployments.
   3. Methodology & Tools: New algorithms, techniques, software, instruments, or experimental frameworks.
   4. Review & Survey: Literature reviews, meta-analyses, systematic reviews, or perspectives.
   5. Other: Education, policy, news, editorials, or papers that do not fit the above.
   6. Irrelevant: Not related to the keyword.

   Task:

   1. Assign the most appropriate Research Type from the list above.
   2. Extract a specific Topic Tag (1-3 words) representing the core subject (e.g., "Deep Learning", "Gene Editing", "Market Analysis").

   Output Format: Return a valid JSON object:

   {"type": "Research Type Name", "topic": "Topic Tag"}

   . Do not output anything else.

3. Generate Final Report (for AI Refinement): Group the articles by their assigned Research Type. For each type that contains articles, generate a summary section.

   System Prompt:

   Act as a comprehensive research summarizer compiling a "Weekly Research Update".

   Input: A list of research papers (Title, Journal, Abstract, Topic Tag) belonging to the Research Type: "{category}".

   Task: Write a concise, engaging summary for this research type.

   • Synthesize: Group papers with similar Topic Tags and summarize their collective contribution.
   • Highlight: Identify the most significant findings or innovations.
   • Tone: Professional, objective, and adapted to the specific domain of the papers (e.g., formal for physics, analytical for social science). Avoid generic "excitement" unless warranted by a major breakthrough.
   • Reference: List the papers with their Titles and Journals.

   Format:

   {Category Name}

   [General Summary Paragraph highlighting key themes]

   Key Updates:

   • [Topic Tag]: [Summary of findings from related papers]. Refs: [Title] (Journal)
   • ...

   (If a paper stands alone, list it individually)

4. Final Output: Combine all sections into a single Markdown document titled "Weekly Research Report: {keywords}". Add a brief "Executive Summary" at the top highlighting the distribution of papers (e.g., "This week saw a focus on Applied Research in [Topic]...").

Quality Rules

• Source: Must use real data returned from the

  pubmed_search.py

  script. Do not hallucinate papers.
• Coverage: Ensure all retrieved and relevant papers are included in the report.
• Tone: Objective, informative, and structured. Avoid overly sensational language.
• Error Handling: If the script returns no results, output "No significant research articles found for '{keywords}' in the last 7 days."