skills.rest
openvetta
openvettaOFFICIAL
Explore all skills by this author →
WebsiteGitHub

evidence-review

Verify claims against concrete evidence and flag unsupported assertions before reporting conclusions.

154|27|Updated Aug 3, 2026
Search Tags:#fact-checking#research#evidence#review#accuracy#claim verification
One-click install
npx skills add https://github.com/openvetta/open-vetta --skill evidence-review-openvetta
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: evidence-review
Source: https://github.com/openvetta/open-vetta/tree/main/packages/coding-agent/examples/runtime-agents/skills/researcher/evidence-review
Command: npx skills add https://github.com/openvetta/open-vetta --skill evidence-review-openvetta

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Research and analysis often mix verified facts with inferences, making it hard to trust reported conclusions. This Skill enforces a disciplined review pass that ties every claim to concrete evidence and explicitly marks anything unsupported. ## Core Features & Use Cases - Claim-by-claim verification: Lists each claim and identifies the specific evidence supporting it. - Unsupported claim flagging: Explicitly marks claims that lack evidence instead of letting them pass as facts. - Inference separation: Prevents inferences from being promoted to verified facts in final reports. - Use Case: After drafting a research summary on market trends, run an evidence review to confirm each statistic is backed by a cited source and flag any speculative statements before publishing. ## Quick Start Review my research summary and check each claim against the available evidence, marking any unsupported claims explicitly.

Frequently Asked Questions about evidence-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I verify claims in a research report before publishing?▼

List each claim in the report, identify the concrete evidence supporting it, and explicitly mark any claim that lacks evidence. This ensures no inference is presented as a verified fact in the final output.

How to separate facts from inferences in AI-generated summaries?▼

Review each statement and trace it back to a specific source or data point. Statements without direct supporting evidence should be labeled as inferences or unsupported claims rather than facts.

When should I use an evidence review step in my workflow?▼

Use it after drafting research summaries, analysis reports, or any document making factual claims, and before sharing conclusions with stakeholders. It is most valuable when accuracy and source traceability matter.

What are the limitations of claim-by-claim evidence checking?▼

It only verifies claims against evidence already available in the context; it cannot fetch new sources or confirm whether cited evidence itself is accurate. External fact-checking may still be needed for critical claims.

Related Skills

Explore More Education & Research Skills→
DojoGenesisDojoGenesis
2

web-research

Query Brave Search and web_fetch to produce source-attributed research summaries.

Community
Advanced
brudexbrudex

auto-research-agent

Compile dated research reports with explicit citations and source notes.

Community
Advanced
paulinett1508-devpaulinett1508-dev

ai-fact-checker

Verify claims against provided source content and flag unsupported statements.

Community
Intermediate
ethicalcapitalethicalcapital

fact-checker

Identify factual claims in text documents and propose inline edits.

Official
Advanced