Learns the author's natural writing voice from past writing samples and applies it as a soft guide during paper drafting. The goal is personalization, not de-AI-ification — the author's voice should come through in the final text, within the boundaries of discipline conventions.
Design boundary: This is NOT a humanizer. We do not aim to evade AI detectors. We aim to produce text that sounds like the author wrote it, because the author's judgment and style are part of scholarly identity.
academic-paper/agents/intake_agent Step 10 (optional)academic-pipeline Material Passport carries the Style Profile across all stagesacademic-paper/agents/draft_writer_agent, deep-research/agents/report_compiler_agentAsk the user:
"Do you have past papers or writing samples you'd like me to learn your style from? Providing 3+ samples helps me match your natural voice. This is optional."
Requirements: - Minimum 3 samples recommended (1-2 samples produce unreliable profiles) - Samples should be the user's own writing (not co-authored sections they didn't write) - Same language as the target paper preferred - Same discipline preferred but not required
Acceptable formats: PDF, DOCX, Markdown, plain text, pasted excerpts
Analyze each sample across 6 dimensions:
{mean: 22, stddev: 8, rhythm: "variable — mixes 8-word punchy sentences with 35-word complex ones"}{mean_sentences: 5, variation: "moderate — 3-7 sentences, shorter in Methods"}{hedging: ["suggests", "appears to", "may"], transitions: ["However", "In contrast", "Yet"], reporting: ["found", "argued", "noted"], formality: "moderate-formal"}{narrative_ratio: 0.4, density: 2.3, placement: "mixed — narrative for key claims, parenthetical for supporting"}{modifier_density: "minimal — lean prose, few adjectives", abstraction: "concrete — prefers specific examples over generalizations"}{shifts: "noticeable — cautious in Methods, increasingly assertive in Discussion, most personal voice in Conclusion"}Combine the 6 dimensions into a Style Profile artifact (see shared/handoff_schemas.md Schema 10).
Report to the user:
"I've analyzed your writing style from [N] samples. Key traits: - [1-sentence summary of most distinctive trait] - [1-sentence summary of second distinctive trait] I'll use this as a soft guide — discipline conventions always take priority."
When the Style Profile is consumed during writing, apply the following priority hierarchy:
Priority 1 (HARD): Discipline conventions
→ Cannot be violated. E.g., if the discipline requires third-person,
the author's preference for first-person is overridden.
Priority 2 (STRONG): Target journal conventions
→ If the user has specified a target journal, its style norms take precedence.
E.g., Nature requires short paragraphs; author's preference for long paragraphs is overridden.
Priority 3 (SOFT): Author's personal style
→ Applied only where it does not conflict with Priority 1 or 2.
E.g., the author's preferred transition words, hedging patterns,
citation integration ratio — these are safe to apply.
When personal style conflicts with discipline or journal norms:
Style conflict: Author prefers passive voice (72% in samples),
but target discipline (Engineering) conventions favor active voice.
→ Using active voice per discipline convention."Note: Your typical use of [trait] differs from [discipline/journal] convention. I've followed the convention, but you can adjust manually if you prefer your style here."
These dimensions rarely conflict with norms and can be applied freely: - Preferred transition words (within academic register) - Hedging word choices - Reporting verb preferences - Citation integration ratio (narrative vs parenthetical) - Modifier density (as long as precision is maintained) - Sentence length variability patterns
These dimensions may conflict with discipline/journal norms: - Voice (active vs passive) — discipline-dependent - Paragraph length — journal-dependent - Person (first vs third) — discipline-dependent - Formality level — journal-dependent
If user provides < 3 samples: generate a partial profile with a warning.
"I have a preliminary style profile from [N] sample(s), but it may not be fully representative. I'll apply it cautiously."
If samples are in a different language than the target paper: extract transferable dimensions only (paragraph structure, citation style, modifier density). Skip vocabulary preferences.
If user indicates samples are co-authored: ask which sections they wrote. Analyze only those sections.
If samples span many years: weight recent samples more heavily (2x weight for samples within 2 years).