At McKinsey, a recommendation was only as good as the research under it. So I learned to do it right: scope the real question, pull from independent sources, triangulate what's actually true, and ground every claim in evidence. This Skill runs that discipline for you — breaking down a question, comparing sources, verifying claims, and writing a structured report with citations you can check. Decisions on evidence, not vibes.
Hard verification gates that catch the claims that get published content deleted — before you hit publish. Five gates reject historical over-generalization, causal claims about tides/weather without primary data, sources that don't cover the claimed scope, stale time-sensitive facts, and one-sided reassurance. Born from real post-publication deletions on a daily knowledge channel. Works on scripts, articles, newsletters, and social posts.
Research Context Management helps agents run long, evidence-heavy research without losing sources, claims, contradictions, or next steps. It builds a durable workspace with a source ledger, batch notes, claim matrix, contradiction register, working brief, and handoff checkpoint, so deep research, audits, market scans, and reports can be reviewed, resumed, and turned into reliable outputs with clearer evidence trails.
Turn web research into verified answers, evidence trails, durable memory proposals, long-form notes, and reusable agent playbooks.
Turn pasted AI evaluation cases into an evidence-bounded triage brief that separates expected and observed behavior, hypotheses, and missing checks without claiming a root cause or fix.
Forces AI agents to prove a task is actually done before declaring it complete. 4-phase gate: instruction reconciliation, evidence anchoring (values only readable from the actual output), self-scoring (A/B/C/D), and reverse witness check. Eliminates "done" declarations that aren't.
Turn local research hypotheses into traceable evidence. Validate daily CSVs, compare predeclared moving-average strategies with chronological holdouts, stress transaction costs, and independently audit results. Includes a Python standard-library lab, visual HTML reports and optional local Vibe-Trading helpers. Analysis and offline simulation only; no access to real user accounts.
Learn from what people actually do, rather than treating praise as demand. Your agent prepares focused research questions and examines supplied interview notes for recent tasks, workarounds, alternatives and concrete commitments. It separates observations, reported experiences and hypotheses, preserves conflicting feedback and proposes the smallest useful next validation. Existing evidence is reused, so the process does not begin with a large questionnaire every time.
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.