BookOS v2.1.0

BookOS v2.1.0

BookOS v2.1.0 makes deep book analysis more source-grounded, complete, and safer to publish.

Highlights

• Verifies a complete supplied source across the beginning, middle, major turning points, and ending.
• Requires a substantive content summary instead of returning only an outline or chapter list.
• Separates source-supported plot events, analysis or interpretation, and external real-world cases.
• Distinguishes PDF viewer pages from printed page numbers and omits unverified locators.
• Uses an exact 17-file allowlist for the complete public archive so unrelated tracked files are not packaged.

Install

• Ordinary download: BookOS-Skill.zip. This stable URL follows the latest published release.
• Codex: ask $skill-installer to install https://github.com/daisymyu-dev/BookOS/tree/main/bookos.
• Versioned provenance asset: BookOS-Skill-v2.1.0.zip (byte-for-byte identical to the stable asset).
• Developers and local knowledge-base users can download BookOS-v2.1.0-public.zip.

Import availability can vary by product, account, and workspace. Codex users can continue using the repository installer route.

Upgrade from v2.0.0

Reinstall or replace the previous Skill folder to use the v2.1 source-grounding rules. Existing private Markdown notes remain local and do not need to be migrated. The optional Python companion reports version 2.1.0.

Privacy

The public assets contain only reviewed project and Skill files. They do not include supplied books, extracted text, screenshots, private notes, generated outputs, local paths, or credentials.

Verification

The release is built only after the full unit-test suite, official Skill validation, archive-member inspection, and privacy scanning pass. SHA-256 digests are also recorded by GitHub for the uploaded assets.

BookOS v2.0.0

A no-code-first BookOS release for turning books and reading notes into structured, memory-friendly knowledge. Includes four Skill modes, an atomic offline knowledge graph, bilingual onboarding, an optional Python companion, privacy-checked deterministic release archives, and a safe anonymous demo.