Inventory conversation claims and categorize each as verified fact or unverified assumption.
npx skills add https://github.com/samcdavid/dotfiles --skill prove-it-samcdavid
Please help me install this Agent Skill. Skill: prove-it Source: https://github.com/samcdavid/dotfiles/tree/main/claude/skills/prove-it Command: npx skills add https://github.com/samcdavid/dotfiles --skill prove-it-samcdavid
Lightweight fact-check of the current conversation. Separates verified facts from unverified assumptions, cites evidence for each, and flags trust debt. Use when findings feel uncertain or before acting on research.
Review the current chat and list each factual claim with supporting evidence, then label any assumptions and flag trust debt.
High-intent search queries and answers about installing and using this skill.
To fact-check a conversation, inventory all factual claims and categorize each as a Verified Fact or Unverified Assumption. Cite precise evidence for verified facts, explain what was assumed for unverified claims, and flag any trust debt.
Trust debt consists of unverifiable or outdated claims within a conversation. You identify it by inventorying all factual claims, separating verified facts with citations from unverified assumptions, and flagging items that cannot be currently verified or need retraction.
To verify factual claims before taking action, inventory the conversation's claims and categorize them as Verified Fact or Unverified Assumption. For each verified fact, cite precise evidence like file paths, command outputs, or documentation.
Yes, you can use this to find evidence citations by inventorying all factual claims from your conversation. It separates verified facts from unverified assumptions, citing precise evidence such as file paths and line numbers or documentation for verified claims.
You should run a fact-check when findings feel uncertain or before acting on research and making decisions. It inventories all factual claims, separates verified facts from unverified assumptions with citations, and flags trust debt.
To retract claims with trust debt, the fact-check identifies the unverifiable or outdated claims and provides specific steps to verify or retract them. It categorizes each claim as a Verified Fact or Unverified Assumption and flags the appropriate trust debt.
bordenetValidate wiki page claims against git history, issue trackers, and meeting transcripts.
ArabelaTsoAnalyze Isabelle, Coq, and Dafny artifacts to report verified and unverified components.
Unson-LLCImplement a six-phase debugging framework that verifies facts before fixing root causes.
ntaksh42Detect and correct errors in AI outputs with automated validation reports.
