This Skill is built on practical expertise in claim-evidence mapping, source authority labeling, quote verification, citation hygiene, recency checks, and hallucination-risk control. It helps AI agents turn research, reports, summaries, briefs, and marketing drafts into auditable claim ledgers where every important claim is matched to a source, date, quote/paraphrase boundary, and confidence label.
Audit AI agent workspaces for authority, secrets, data flow, memory, supply-chain, and evidence risks—locally and without executing target code.
Paste AI evaluation cases and get an evidence-bounded triage brief: expected vs. observed behavior, ranked hypotheses, and the missing checks to run next — so you can decide what to investigate first without claiming a root cause or fix.
Structures field notes, photo descriptions, and claim details into a clean claims file with evidence inventory, completeness view, and dual internal + summary versions.
Evidence-based technical SEO audit of a whole website with a prioritized fix list.
Paste decision records and meeting notes. Get a source-linked consistency report: a conflict table that names the conflicting choices, a judgment on which statements actually contradict, and the unresolved questions to settle before you act.
Keep and check the evidence behind a research answer. Your AI agent finds sources and saves their text; a bundled offline Python script verifies every quote against the saved text, treats copied articles as one origin, flags conflicting numbers, old sources and single-origin findings, and seals the ledger with a hash. Exports a folder ClaimLock can read. Download package: you bring your own AI agent and Python 3; no model included.
Check AI-generated claims against your approved Fact Sheet and flag changed quantities, expanded scope, dropped attribution, and unsupported mappings — deterministic, local, and no AI guessing.
Know if a betting record is luck or skill. Paste any betting history — get a full statistical audit: ROI with 95% confidence interval, the probability a losing bettor would show the same record, drawdown analysis, risk of ruin, and curation red flags. Retrospective analysis only — never predictions.