Skip to content

v4.0.0 · End-to-end humanities writing assistant

Choose a tag to compare

View all tags
@tizzy916 tizzy916 released this 19 May 03:38
v4.0.0
50fbaec

Major strategic repositioning + capability expansion.

This release removes the "humanities-side companion to academic-research-skills" framing. v4.0 is an independent end-to-end writing assistant for humanities scholars covering the full lifecycle of a humanities paper, from research-question sharpening to AI-use disclosure for submission.

Four new modes (pre-writing + submission-end coverage)

  • Mode H · Research-question sharpening (Socratic) — humanities-specific question taxonomy (re-reading / re-construction / intervention), so-what test, interlocutor identification. Output: _writing-config/research-question.md.
  • Mode I · Literature mapping — organizes references the author has already read into a camp-and-debate map. Hard rule: does not search literature for the author. Optional scholar-wendao integration. Output: _writing-config/literature-map.md.
  • Mode J · Plan-only outlining — discipline-aware, pulls discipline-specific standard arcs. Hard constraint: does NOT write paragraphs. Output: _writing-config/outline.md.
  • Mode K · AI-use disclosure (humanities-journal-specific) — 4-tier scheme + 3 template options. Audits interaction-log + revision-log. Warns when tier exceeds typical humanities journal policy. Output: _meta/AI-use-statement.md.

Enhanced existing modes

  • Mode D calibration: 1-5 reviewer intensity levels (gentle reader → adversarial committee member). Concession Threshold tightens at Level 5.
  • Mode D methodology-focus sub-mode: attacks methodology rather than content. Discipline-aware vulnerability set.
  • Mode F.coach sub-mode: skill withholds revision until author answers 3-5 diagnostic questions. Trains author to internalize four-layer critique.

Citation toolchain (new)

  • scripts/citation-format-convert.py: BibTeX → Chicago / MLA / APA / GB/T 7714 converter. Zero deps. Sanity-tested in all four styles.
  • scripts/citation-verify.py: in-prose citations verified against Crossref API. Three verdicts (FOUND / FUZZY_MATCH / NOT_FOUND). Rate-limited 1 req/sec.

Removed

  • ## Companion: academic-research-skills section. ARS demoted from designated companion to one of several lateral tools.
  • ARS row removed from Comparison-with-adjacent-tools table.

See CHANGELOG.md for full history.

Assets 2
Loading