$ loading_
Evaluate manuscript retrieval and AI grounding with deterministic evidence and coverage scoring.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "LongBook Verifier" yet — see the docs or source repo.
Use LongBook Verifier to validate this long-form novel manuscript and its AI-generated summary: chunk and index the manuscript, retrieve evidence for each key claim in the summary, and output grounding coverage, unsupported sentences, and traceable source passages.
Returns chunking and retrieval results, evidence mapping for each summary claim, an overall coverage score, and a list of unsupported content.
Use LongBook Verifier to check the answer quality of a Q&A system on this long-form nonfiction manuscript: retrieve supporting evidence for each answer, determine whether it is faithful to the source, and report evidence coverage, omissions, and possible hallucinations.
Outputs supporting evidence for each answer, a faithfulness judgment, coverage scores, and annotations for omissions and possible hallucinations.
Use LongBook Verifier to compare two retrieval methods on this manuscript: run chunking, indexing, and retrieval for each method, then evaluate evidence hits, coverage, and consistency on the same question set, and provide quantified comparison results.
Generates retrieval hit rates, evidence coverage scores, consistency comparisons for both methods, and a recommendation on which approach to use.
Extract, verify, cache, and search factual claims from text.
Validate AI-generated code with browser tests, evidence capture, and smart diagnostics.
Validate AI outputs against contracts and auto-repair noncompliant results.
Conduct source-grounded web research and verify claims against original sources.
Identify antiquarian first editions and manage New Mexico book donation logistics.
Manage a book library with natural language for search and CRUD tasks.