Add automated paper summarization for active recall issues

[Copilot]

Implements automated PDF summarization for arXiv papers linked in active recall issues. Summaries are triggered via comment (@learning-review-bot summary) or automatically after 3 days.

Implementation

Workflow (.github/workflows/paper-summary.yml)

• Triggers: issue comment, daily schedule (7 AM EST / 12 PM UTC), workflow_dispatch
• Finds issues with daily-learning + llm labels >3 days old without summaries
• Rate limits to 3 summaries/day, prevents duplicates
• Extracts arXiv URLs supporting both http and https protocols: https?://arxiv\.org/abs/[0-9]{4}\.[0-9]{4,5}(v[0-9]+)?
• Includes arXiv availability check before processing
• Posts failure notifications to issues when comment-triggered workflows fail

PDF Processing (scripts/summarize-paper.js)

• Converts arXiv abstract URLs (/abs/) to PDF URLs (/pdf/)
• Downloads PDF with redirect protection (MAX_REDIRECTS=5) and 60-second timeout
• Extracts text using cached pdf-parse module (dynamic import for CommonJS compatibility)
• Truncates to 50K chars, generates summary via GPT-4o-mini
• Outputs markdown-formatted summary to GitHub comment

Usage

# Manual trigger
@learning-review-bot summary

# Local testing
node scripts/summarize-paper.js http://arxiv.org/abs/2601.00698v1

Summary includes: research question, approach, key findings, significance, limitations.

Dependencies

• pdf-parse@^1.1.1 for PDF text extraction

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Workflow to Summarize Research Paper</issue_title>
<issue_description>Have a GitHub Action workflow that will review the PDF assigned to the active recall issue. After reviewing, a comment will be added to the issue with the summary of the paper. This action will happen if summoned by a comment "@learning-review-bot summary" or if the active recall is past 3 days. The PDF will be at the same URL that the source paper is, except that the URL will have pdf rather than abs in it. For example, the source link is http://arxiv.org/abs/2601.00698v1 and the pdf for the source is http://arxiv.org/pdf/2601.00698v1. The model will read from the PDF and come up with a high level summary that includes the key points from the paper.</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Workflow to Summarize Research Paper #157

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[coderabbitai]

Important

Review skipped

Bot user detected.

To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[pertrai1]

@copilot please make updates based on review

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 948bf2a0cc

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".