Routing Rules — Full Detail

Extended routing guidance for latex-paper-en. The SKILL.md keeps only the core rules; this file preserves the full decision notes.

Module inference

• Infer the module from the user request before asking follow-up questions. Ask for the module only when two or more modules are equally plausible after keyword routing.
• If the user asks for 2-3 compatible checks in one turn, run them sequentially instead of forcing a single-module reply.
• Execution order when multiple modules are needed: compile -> bibliography -> format -> figures / tables / caption / pseudocode -> grammar / sentences / deai -> logic / literature / experiment / abstract -> section-writing -> title / expression / translation / adapt.
• When applying multiple polish passes to the same prose, work coarse-to-fine — argument/logic -> sentence structure -> lexical/formatting — and do not reverse it; see references/modules/workflow.md.

Choosing between adjacent modules

• Prefer logic for cross-section alignment requests (abstract vs introduction vs conclusion), introduction funnel issues, or contribution drift; prefer literature only when the problem is specifically about Related Work organization, comparison, or gap derivation.
• Prefer section-writing when the user asks to draft, rewrite, restructure, or reviewer-polish a specific section, or asks for paragraph roles, mini-outlines, reverse outlines, or claim-evidence maps. Prefer the diagnostic modules first when the user asks to check whether something is wrong.
• Keep experiment for results, discussion, baseline, ablation, significance, limitation, and conclusion-completeness concerns even if the user phrases them as "logic" problems.

Special flags and loading rules

• For section-writing, load references/modules/section-writing.md, then exactly one active section guide from references/writing/section-writing/ unless the user also asks for flow or self-review.
• For whole-paper motivation/red-thread questions ("does every introduction promise get tested and resolved?"), run logic with --motivation-thread; it appends a read-only Promise Map + Closure Map heuristic and leaves default logic output unchanged.
• For graded de-AI / AIGC-dimension analysis, run deai with --tier light|medium|heavy; it scales thresholds, adds a D1 sentence-length check, and labels findings by dimension (D1-D5). Omitting --tier keeps the default output.

Failure handling

• When a script fails, stop the current module, report the exact command plus exit code, and recommend the next smallest useful fallback instead of silently switching modules.

Output contract details

• For literature, default to diagnosis + rewrite blueprint first; only produce paragraph-level rewriting when the user explicitly asks for prose.
• For section-writing, return a section objective, compact outline, paragraph roles, rewrite blueprint or prose proposal, claim-evidence map, and self-review checklist. Mark missing evidence instead of filling it.

Safety rationale (full text)

• Don't invent citations, metrics, baselines, or experimental results — fabricated evidence is harder to retract once the user trusts it than a clearly flagged gap.
• Leave \cite{}, \ref{}, \label{}, custom macros, and math environments untouched by default — a stray edit there is far harder to spot in a diff than a prose edit, and breaks compilation silently.
• Treat generated prose as proposals, not commits — keep source-preserving checks separate from rewriting so the user can validate each step.
• Do not enable online bibliography checks unless the user explicitly asks for external verification or confirms that citation metadata may be sent to third-party APIs.
• The deai module improves readability; it is not a detector-evasion tool and does not remove a disclosure obligation. If an LLM had a non-trivial role in the paper, remind the user to disclose it per the target venue's policy (references/venues/ai-disclosure.md has the per-venue matrix — some venues require disclosure in the cover letter, a dedicated section, or the checklist).