npx claudepluginhub buriedsignals/skills --plugin spotlightThis skill uses the workspace's default tool permissions.
When a source is behind a paywall or access barrier, work through this hierarchy before marking it inaccessible. A paywalled source should not default to `low` confidence — it should default to an access attempt.
Assigns reputation tiers to sources, cross-references claims for verification, detects biases, and standardizes citations with metadata.
Finds primary sources for claims (arXiv papers, official docs, RFCs, textbooks) enforcing tiered hierarchy. Records URL, title, verifying passage/result.
Searches academic literature via arXiv, Semantic Scholar, and open-access sources. Fetches and parses PDFs for abstracts, key findings, methodology, and citations. Use for research, literature reviews, or formal citations.
Share bugs, ideas, or general feedback.
When a source is behind a paywall or access barrier, work through this hierarchy before marking it inaccessible. A paywalled source should not default to low confidence — it should default to an access attempt.
Try each step in order. Stop as soon as you have the full text.
Many paywalled articles have freely distributed copies:
"{exact title}" filetype:pdf
"{exact title}" site:{author-institution}.edu
"{exact title}" site:researchgate.net
"{exact title}" site:academia.edu
curl "https://api.unpaywall.org/v2/{DOI}?email=research@buriedsignals.com"
Parse the response for best_oa_location.url — if present, it's the legal open-access copy. Also check oa_locations[] for mirrors.
curl "https://api.core.ac.uk/v3/search/works?q={query}&limit=5" \
-H "Authorization: Bearer {CORE_API_KEY}"
Check results[].downloadUrl for PDF links.
curl "https://api.semanticscholar.org/graph/v1/paper/search?query={query}&fields=title,openAccessPdf,externalIds&limit=5"
Check openAccessPdf.url.
Check if an archived version is accessible before the paywall went up:
# Check Wayback Machine
curl "http://web.archive.org/cdx/search/cdx?url={URL}&output=json&limit=3&fl=timestamp,statuscode"
# Check Archive.today
curl -s "https://archive.ph/newest/{URL}"
Early snapshots often predate paywall implementation.
Search the title in Google Scholar. Look for [PDF] links next to results — these are often author-posted copies on institutional servers.
For publications with free article limits (e.g., Financial Times, Bloomberg):
⌘-Shift-R or View → Show ReaderF9 or the reader icon in the address barWorks only on metered paywalls, not hard paywalls.
Email the author directly. Most authors are willing to share their work:
Subject: Research access request — [{Title}]
I'm a journalist investigating [{topic}] and would like to access your paper "{Title}" (published {year}). Would you be able to share a copy? I'm happy to credit your work in my reporting.
Add access_method to every source entry in findings.json and fact-check.json:
{
"url": "https://example.com/paywalled-article",
"type": "news",
"accessed": "2026-03-15T14:20:00Z",
"access_method": "archive_copy",
"access_notes": "Retrieved via Wayback Machine snapshot from 2024-11-02"
}
Valid values for access_method:
full_text — Complete article accessed directlyopen_access — Legal open-access copy found via Unpaywall/CORE/authorarchive_copy — Accessed via Wayback Machine or Archive.todayabstract_only — Only abstract/summary availableinaccessible — Could not access after full hierarchy attempt| access_method | Confidence cap |
|---|---|
full_text / open_access | No cap |
archive_copy | No cap (but note snapshot date) |
abstract_only | medium at best — abstract may omit key findings |
inaccessible | low — cite the source but flag that content was not verified |
A finding that rests on an inaccessible source must say so explicitly in confidence_rationale. Do not inflate confidence because the source sounds authoritative if you have not read it.
inaccessible.access_notes so the next cycle doesn't repeat the effort.Adapted from claude-skills-journalism by Jay Amditis, released under MIT License.