abstract-searcher

What This Skill Does

The abstract-searcher is a highly specialized OpenClaw agent skill designed to streamline the academic writing and bibliography management process. It automates the tedious task of populating BibTeX files with missing metadata. By leveraging a multi-tiered approach, it first queries major academic databases including arXiv, Semantic Scholar, CrossRef, and OpenAlex. If these automated API queries fail to yield a result, the skill gracefully transitions to an advanced browser automation mode. In this mode, it utilizes your local Chrome profile to navigate Google Scholar, mimic human search behavior, and extract abstract text directly from publisher websites like ScienceDirect, Springer, or ACL Anthology. This ensures high coverage for papers that are either behind paywalls or missing from public API endpoints.

Installation

To add this capability to your OpenClaw agent, use the official installation command within your agent terminal:

clawhub install openclaw/skills/skills/easonc13/abstract-searcher

Ensure that your environment has access to the Chrome browser and that you have installed any necessary dependencies required by the script.

Use Cases

1. Automated Literature Review: Quickly aggregate missing metadata for a large collection of downloaded PDFs to organize your Zotero or BibTeX library.
2. Academic Publication Prep: Automatically fix 'stub' citations in a LaTeX project before submission to ensure your references follow standard formatting.
3. Cross-Platform Synchronization: Sync bibliographic data between different platforms where abstracts may have been stripped during export.

Example Prompts

1. "OpenClaw, please run the abstract-searcher on references.bib located in my documents folder and overwrite the file with the updated abstracts."
2. "I have a list of new papers in research_notes.bib; use the abstract-searcher to fill in all missing abstracts, and use Chrome if the APIs don't return anything."
3. "Check for missing abstracts in my main library.bib file and generate a report of any papers you couldn't find, even after using the browser fallback."

Tips & Limitations

• Browser Integration: Always ensure your Chrome profile is synced and active. If you encounter errors, verify that you have logged into your institutional account via the browser so you have access to paid journals.
• Rate Limiting: The script respects API rate limits with a built-in 2-second delay. Do not force-terminate the process, or you may hit temporary IP bans from services like CrossRef.
• Data Integrity: Always perform a quick sanity check on the outputted BibTeX. While the tool is highly accurate, verifying the title match before final submission to a journal is recommended.
• Resource Usage: Running the browser automation mode is resource-intensive; ensure your workstation is not overloaded before initiating a full-library scan.