Abstract Fetcher

Fetch paper abstracts from multiple open access sources.

Definition

• •ID: abstract-fetcher
• •Type: Command
• •Author: Research Skills
• •Description: Retrieve abstracts from Semantic Scholar, OpenAlex, and web search

Commands

Single DOI

/abstract-fetcher fetch "10.1145/1234567"

Get abstract for a single paper.

Batch Processing

/abstract-fetcher batch input.json output.json

Process multiple DOIs from JSON file.

Check Status

/abstract-fetcher status results.json

Check success/failure statistics for batch run.

Requirements

• •Python 3.10+
• •requests>=2.31.0
• •No API keys needed (uses public APIs)

Setup

1. •

   Install as part of skills-research:

   bash

   cd ~/.claude/skills/skills-research/literature-search/abstract-fetcher
   

2. •

   Configure (optional):

   bash

   cp .env.example .env
   

Environment Variables

• •ABSTRACT_FETCHER_DELAY - Rate limiting delay (optional, defaults to 0.5s)

Sources (Cascading Order)

1. •Semantic Scholar API - Most complete abstracts
2. •OpenAlex API - Fallback with good coverage
3. •Web Search - Last resort for rare papers

Examples

/abstract-fetcher fetch "10.1145/3706598"
/abstract-fetcher batch dois.json abstracts.json
/abstract-fetcher status abstracts.json

Input Format (for batch)

[
  {"doi": "10.1145/3706598"},
  {"doi": "10.1145/3145816"}
]

Output Format

[
  {
    "doi": "10.1145/3706598",
    "abstract": "Abstract text...",
    "source": "semantic_scholar"
  }
]

Notes

• •Automatic rate limiting to avoid hitting API limits
• •Gracefully handles missing abstracts
• •Batch processing with progress reporting