Technical format: Skill
Claim Review
Review a supplied claim, its scope and counterevidence before a person decides what to share.
What Claim Review does
A focused evidence-ledger method with an optional structural validator and not-reviewed worksheet generator.
- For
- Researchers, authors and editors checking a claim before sharing.
- Bring
- Permitted supplied claims, evidence excerpts and locators, context, intended reader and decision.
- Action
- Preserve original claims and context, inspect support/contradiction/unknowns, retain all supplied evidence references and propose scoped revisions.
- Output
- A human-owned claim ledger, uncertainty and a clear revise/defer/sharing decision point.
Runtime and prerequisites
The method runs in your existing AI agent. Its optional structural validator and not-reviewed worksheet generator use Python 3.11+ and the standard library; no source search or semantic classifier is bundled.
- An Agent Skills-capable assistant or direct use of the method; Python3.11+ only for the optional helper. No account, sibling capability or package installation.
Works on its own: No other capability, account or package is required. The helper is optional; semantic review remains a human-owned decision.
Natural stop: Return the ledger and scoped revise/defer decision, then hand control back. Any practice is opt-in and stops at user stop, agreed cap, missing input or a failed/unknown critical gate.
How the work moves
Example output
A fictional claim ledger preserving the original claim, source references, counterevidence and open questions.
Fictional illustration, not a result from your task or measured user evidence.
Limitations
Structure, reference coverage and exact input binding only. No source fetching, quote verification, truth certification, publication approval, independent human usefulness or measured coaching improvement.
Explore by kind of work
Another part of the work.
Back to the collection
Related tools & research
Sources to compare or explore alongside this capability. A listing does not imply an integration, partnership or endorsement.
Research & guidance 2 sources
Cochrane Handbook: Collecting data
Primary data-extraction guidance makes participants, setting, outcomes, linked reports and discrepancies explicit.
Difference: Claim Review is a general supplied-packet review, not a systematic review or clinical assessment. No Cochrane text, form or code is copied.
Reviewed
W3C PROV overview
Provenance concepts distinguish a reviewed claim, its source and the activity producing a review.
Difference: Local IDs and a canonical packet digest are used; no PROV serialization, ontology, implementation code or conformance claim. The linked PROV-DM 2013 Recommendation was checked as a related standard after implementation.
Reviewed
Commercial tools 1 source
Scite
A commercial research service offers support/contradiction signals from citation context.
Difference: Claim Review reviews supplied evidence locally; it implements no search, citation index, Scite API, classifier or account. Vendor outcome and coverage claims are not adopted.
Reviewed