Create English paper digests with AI
Build an English research digest and local reading queue with verified paper identities, evidence-rich cards and a history that avoids repeated recommendations.
Create English paper digests with AI around an agreed research question and selection policy. The paper-research-daily skill combines an agent’s reading and judgement with local helpers for retrieval, identity checks, history and report assembly.
The Python tools do not autonomously understand papers or write scholarly reviews. Documented output languages are English and Chinese. Scores remain reasoned editorial assessments, rather than calibrated probabilities, and insufficient evidence must be reported instead of lowering standards to fill a daily quota.
What it does
- Build a question-led reading queue
- Prepare detailed evidence-rich reading cards
- Verify publication status and downloaded versions
- Avoid repeated recommendations across versions
- Retain dated reports, sources and selection history
- Save incomplete work without claiming a finished digest
Install and use
- Load the complete package in an agent with web retrieval, file access and command execution. Prepare Python 3.11+ with the declared dependencies and a writable local library, identifying any existing papers first.
- Ask: “Prepare an English digest on learning from limited data. Agree sources, quantity and selection standards with me before collecting papers, and check my existing library for duplicates.”
- Review the profile and output language. Discovery, official publication verification, lawful full-text access and reading depth are separate checks; a DOI alone does not establish peer review.
- Inspect the cards, version identifiers, links and saved files. Keep shortfalls explicit; uploading, sending the library or installing recurrence requires separate instructions.
FAQ
Source
Made by paper-research-daily contributors. Original skill: paper-research-daily, published on GitHub under the MIT licence.
View on GitHub