| name | book-author |
| description | Render surviving research findings into a book chapter through the typed findings-to-artifact contract. An OPTIONAL channel pack (enable the `book` pack to use it), supporting the full genre range. Use this skill when the user wants to author a book, write a chapter from the research, turn the corpus into long-form, or build a manuscript. Triggers on 'write a book', 'author a chapter', 'book from research', 'turn this into a chapter', 'long-form from the corpus', 'manuscript'. |
| version | 0.4.3 |
| argument-hint | [--topic <id>] [--genre <genre>] [--chapter <n>] [--out <path.md>] |
| allowed-tools | Read, Bash, Edit, Glob, Grep |
book-author — research → book chapter
Book is an optional channel pack (SPEC §6d): enable the bundled book channel
pack to author long-form — it is not an always-on core output (blog is the
first-class published channel). It renders a chapter from the surviving findings
through the same typed findings→artifact contract (schemas/artifact.schema.json)
the blog output uses — so a research corpus drives both a post and a book from one
synthesis, and
the citation gates run uniformly across both.
The book pipeline supports its full genre range (technical / children's / fiction
/ history / non-fiction) via the genre passed to the synthesizer, and the chapter
carries front/endnote matter. The citation-leak gate is the manuscript invariant:
a chapter must read as if written from public primary sources alone.
Pipeline
-
Resolve the finding set for the chapter — the topic's surviving findings
(verdict ≠ falsified).
-
Synthesize the artifact (the report-synthesizer substrate):
scripts/synthesize-artifact.sh reports/<topic>/findings <genre> reports/<topic>/artifact.json
-
Render the chapter from the same artifact the blog uses. Every output
lives under reports/<topic>/, never a top-level book/ directory — the
book's own chapters/front-matter/appendices structure nests one level
deeper instead:
scripts/render-artifact.sh reports/<topic>/artifact.json book reports/<topic>/book/chapters/<n>.md
<n> is a numeric chapter index for a single manuscript authored this way,
chapter by chapter. The research-deliverables.js multi-genre routing
workflow drives this same render step differently: since a numeric index
doesn't map onto "N independent genre-shaped chapters" rendered in one
run, it uses the genre string itself as the chapter slug instead
(reports/<topic>/book/chapters/<genre>.md) — a deliberate override for
that caller, not a different convention this skill itself follows.
-
Gate the manuscript. The bundled check-citation-leak.sh hook guards
reports/*/book/chapters/*.md (and the matching appendices//front-matter/
paths): no finding ids, urn:mif: ids, or reports/<slug>/ paths may appear.
Re-author any flagged passage from the primary source.
Non-negotiables
- Never include a falsified or quarantined finding — the synthesizer filters
them.
- Manuscript prose carries no internal-research references. The chapter cites
only the public sources in its endnotes.
- The chapter is a projection of the artifact; change the findings or synthesis,
not the rendered Markdown in place.
book is a MIF-exempt channel (declared mifExempt: true in
harness.config.json outputs[]): its public prose is orthogonal to MIF, so the
MIF Level-3 source of truth lives in the generic report channel
(reports/<topic>/<slug>.md), not the chapter. The MIF I/O conformance gate skips
and logs the chapter; it does not require an L3 projection of it.
