academic

Files (21)

• .gitignore
• SKILL.md
• __init__.py
• SKILL.md · arxiv/
• arxiv.py · arxiv/scripts/
• SKILL.md · crossref/
• crossref.py · crossref/scripts/
• SKILL.md · ieee/
• ieee.py · ieee/scripts/
• SKILL.md · openalex/
• openalex.py · openalex/scripts/
• SKILL.md · references/
• requirements.txt
• __init__.py · scripts/
• common.py · scripts/
• SKILL.md · semantic_scholar/
• semantic_scholar.py · semantic_scholar/scripts/
• SKILL.md · systematic/
• systematic.py · systematic/scripts/
• SKILL.md · unpaywall/
• unpaywall.py · unpaywall/scripts/

SKILL.md (first lines)

## Academic Provider Directory

Use provider-specific skills for the best fit:

- `academic/arxiv` — preprints and open-access papers on arXiv
- `academic/crossref` — DOI metadata and publisher-level bibliography
- `academic/semantic_scholar` — citation-aware discovery and abstracts
- `academic/openalex` — broad metadata, institutions, and funding context
- `academic/ieee` — IEEE Xplore coverage for applied engineering and electronics
- `academic/unpaywall` — legal open-access PDFs for known DOIs

For structured literature synthesis, use `academic/systematic`.
For reference list building and export, use `academic/references`.

## When to use this skill

- You want a high-level academic search strategy and provider selection guide.
- You need to decide which database is most effective for a research question.
- You want to combine provider-specific results into a single systematic review workflow.

## Typical workflow

1. Identify the research need.
2. Choose providers that match the need:
   - start with `academic/semantic_scholar` for broad search and citation context,
   - use `academic/openalex` for author/institution/funding metadata,
   - use `academic/arxiv` for preprints and open versions,
   - use `academic/crossref` for DOI metadata, publisher, and bibliography details,
   - use `academic/ieee` for IEEE-focused applied research,
   - use `academic/unpaywall` to resolve PDFs by DOI.
3. Collect candidate papers.
4. Deduplicate and screen by title/abstract.
5. Use `academic/systematic` to formalize inclusion and synthesis, or `academic/references` to build an exportable reference list.

## CLI Quick Reference

The systematic review engine can be run from the command line:

```bash
# From the project root:
.venv/bin/python skills/academic/systematic/scripts/systematic.py review "transformer attention" --limit 20 --from-year 2022
.venv/bin/python skills/academic/systematic/scripts/systematic.py sources "LLM alignment"
.venv/bin/python skills/academic/systematic/scripts/systematic.py metrics results.jsonl
```

## Example questions

- "Find the most relevant open-access arXiv and IEEE papers on demand forecasting."
- "Search for citations and abstracts for papers on distribution network optimization."
- "Resolve the open-access PDF for DOI 10.1145/xxxxxxx."
- "Build a deduplicated reference list from these search results."

## Related skills

- `academic/systematic` — Multi-source search and PRISMA screening
- `academic/references` — Reference list building and export
- `academic/arxiv` — Preprint search
- `academic/crossref` — DOI metadata
- `academic/semantic_scholar` — Citation-aware discovery
- `academic/openalex` — Broad scholarly metadata
- `academic/ieee` — IEEE Xplore search
- `academic/unpaywall` — Open-access PDF resolution

From lseman/logician, MIT.

More from lseman/logician