Install
$ agentstack add skill-123trump-latest-paper-search-latest-paper-search ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
Latest Paper Search
Search for papers newly published or newly indexed today, this week, or since a given year, then produce a lightweight report. The skill is self-contained and agent-agnostic: it can be installed under Codex, Claude Code, OpenClaw, project-local skill folders, or run directly from its Python scripts.
Workflow
- Choose the time window:
today: search the past 2 days to buffer indexing lag.week: search the past 10 days to buffer indexing lag.--since-year YEAR: search from January 1 of that year through today.
- Route sources:
- Use OpenAlex only when
OPENALEX_API_KEYis set. - Use arXiv for AI, computer vision, SAR, image processing, foundation model, deep learning, and preprint-heavy topics.
- Use Crossref for DOI and publisher metadata discovery.
- Use Semantic Scholar enrichment only when
SEMANTIC_SCHOLAR_API_KEYis set. - Use SerpAPI Google Scholar only when
SERPAPI_API_KEYis set; treat it as a supplemental discovery source.
- Run the latest search script from the user's working directory:
python /scripts/latest_search.py "" --window week
Useful options:
python /scripts/latest_search.py "" --window today --limit 30
python /scripts/latest_search.py "" --sources openalex,arxiv,serpapi
python /scripts/latest_search.py "" --sources serpapi --since-year 2026
python /scripts/latest_search.py "" --language zh --output-dir
- Report the generated
report.mdpath and summarize the top findings briefly.
Output
Each run writes to the current working directory unless --output-dir is provided:
paper-latest-results/__/
report.md
papers.csv
papers.json
kg.json
run_log.json
raw/
The report must stay lightweight:
- Topic, time window, and sources used
- Raw hit count and deduplicated paper count
- Chinese or English papers only; filter out papers whose titles are dominated by other writing systems
- Top 3-5 papers worth reading with short reasons
- Full paper list with title, authors, keywords, abstract, OA status, PDF/download link, DOI/page link, and source trace
Deduplication
Use this skill's scripts/deduplicate_latest.py for deduplication and field merging. Do not replace it with title-only merging.
Read references/deduplication.md before changing dedup behavior.
References
references/remote_sensing_sources.md: remote sensing routing and venue/source guidance.references/serpapi_google_scholar.md: SerpAPI setup, fields, and limitations.references/deduplication.md: self-contained deduplication and merge behavior.references/report_format.md: report format and ranking rules.references/agent_portability.md: how to install and run the skill across Codex, Claude Code, OpenClaw, and direct CLI usage.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: 123trump
- Source: 123trump/latest-paper-search
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.