Licensed, citable retrieval for books, papers, and deep reference material — built for AI. Bring your owned or institution-licensed sources into your AI workflows with clean citations, clear usage rules, and an upgrade path to paying rights holders.
AI workflows break down where the best information lives. Books and paid PDFs are hard to search well. Citations are messy across editions and formats. Rights are unclear. Teams rebuild the same pipelines in a dozen incompatible ways.
A content and retrieval standard for long-form sources. BCP defines how books, papers, and manuals are described, accessed, cited, and metered — so any tool can retrieve the right passages without treating content like a free-for-all.
What the work is, which edition, how it's structured. Stable work and edition identifiers.
Stable references to passages — not fragile page numbers. Section, paragraph, and text hash.
What's allowed: quote length, display rules, caching limits. Rights travel with the content.
Optional usage metering so creators and publishers can get paid. Micropayments, subscriptions, or one-time access.
Four standard tools — delivered over MCP or any transport.
query, filters → passage handles
handle, max_chars → text + citation + receipt
handle, style → formatted citation
work_id → allowed uses + pricing
search("how repetition creates false beliefs", { author: "Kahneman" })
“A reliable way to make people believe in falsehoods is frequent repetition, because familiarity is not easily distinguished from truth.”
verbatim allowed / max 500 chars
384 chars / $0.0029
Plug books and papers into agents and RAG with clean, stable citations. Less glue code, fewer one-off pipelines, better results.
Build internal knowledge systems over finance, science, and engineering references — with provenance and audit trails.
New revenue channel into AI retrieval. Policy controls, usage receipts, and a real path to getting paid for your work.
Bring-your-own-access with existing subscriptions. Compliance, audit logs, and org-level controls.
We're building the spec in the open. Drop your email and we'll keep you posted as things take shape.