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.
markdown／text ファイル末尾の KNOWLEDGE_ENTRY_V1 ブロックを抽出し、ジャンル別 xlsx 台帳の該当シート末尾へ自動追記します。ID重複検出（冪等）＋非破壊書込（既定は新ファイル出力・--in-place で上書き）。AIセッション横断で研究ノート／チャット要約／ビジネスアイデアを蓄積する knowledge worker 向け。ローカル処理のみ・LLM呼び出しゼロ・ネットワーク不要。
Extracts critical dates, financial terms, and key obligations from leases and produces a clean structured abstract with gap detection.
전자발급 토지대장·등기사항증명서·지적도를 비교해 지번·면적·소유자·권리관계의 불일치와 매수 위험요소를 찾아내고, 계약 전 확인해야 할 사항과 행동 순서를 안내합니다.
Your agent or chatbot keeps getting document content wrong, citing passages that don't exist, or saying "I can't find it" about things plainly in the docs. This reframes it as a retrieval failure, not a prompt bug: corpus profile, chunking, embedding and vector-DB index choice, retrieval strategy (top-k, hybrid BM25 + dense, rerank, query rewrite), context assembly with citations, and a golden Q&A set scored on recall@k / MRR / faithfulness.
Most AI research sounds confident and quietly makes up its citations. This does the opposite: paste your question (and any sources you have) and get a structured brief — summary, findings, synthesis, limitations, recommendations — that reasons from solid knowledge and the sources YOU provide, and never fabricates a citation. Anything it can't ground is marked [UNVERIFIED], so you know what's solid and what to check. For founders, analysts, and anyone who needs a brief that survives scrutiny.