screenplay-prep

Read-only. Lints the project's character cues and names for consistency — dialogue-cue spelling/casing, dialogue-extension grammar (V.O./O.S./CONT'D), first-appearance CAPS introductions, and alias/name drift (read the intentional alias systems from PROJECT_PROFILE §5) — and reconciles every speaking cue against the cast registry (PROJECT_PROFILE §0 `cast_registry`). Flags cue variants, unintroduced characters, stray name spellings, and cues with no casting row. Use after a dialogue or character pass, before re-deriving the casting docs, or whenever a name may have drifted.

Read-only pre-flight check before a commit. Verifies the exported PDF is in sync with the .highland bundle (the thing the pre-commit hook blocks on), .script-state is current, the working tree is sane, and recommends the fact/fidelity audits. Returns a clear GO / NO-GO with the exact blocking reason. Use before committing a script change or a doc batch.

Read-only. Reads the project's screenplay and tracks story continuity across all scenes — time/chronology logic, character knowledge ("who knows what, when"), prop/object setups and payoffs, character physical state, geography, and name/age consistency. Reports continuity errors and unfired setups with scene-number citations. Use after a script revision or a table read, or when reworking a sequence.

Read-only. When an assumption changes (page count, shoot days, a casting/schedule decision), recomputes every downstream number — runtime, shoot-day totals, budget stacks — and shows the arithmetic. Proposes corrections with the math visible; never silently rewrites. Use after a script revision or a production decision that ripples into the numbers.

Read-only. Audits the project's screenplay dialogue for distinct character voice, flagging interchangeable lines (could be swapped between characters unnoticed), on-the-nose exposition, and voice drift for each principal. Reports examples with scene numbers. Use after a dialogue pass or table read, or when a character's voice feels flat.

Run the read-only screenplay audit suite against the current project. Reads PROJECT_PROFILE.md §0 for the per-screenplay parameters, then dispatches the relevant auditor agents (continuity, motif, slugline, cast-cue, fact-drift, fidelity, structure, dialogue) and summarizes findings by severity. Use after a script revision, before circulating a draft, or to spot-check the derived production-doc corpus.

Build a polished, send-ready cast report (landscape PDF + full-width DOCX) from a cast CSV — e.g. a scene-scoped cast list headed to a casting service. Wraps scripts/make_cast_report.py: 10pt, bold headers, each role's row kept whole across pages, descriptions backfilled from casting_editorial.json. Use when the writer needs a formatted casting submission/handout, not the raw CSV.

Re-convert a Final Draft .fdx to the project's Fountain text_mirror after the writer edits the .fdx. Wraps scripts/fdx_to_fountain.py — reads source.fdx + source.text_mirror from PROJECT_PROFILE §0, regenerates the .fountain the engine reads, and reports the delta. Use after every .fdx edit on a Final Draft project (the FD analogue of :reconcile's text-mirror refresh).

Refresh the project's HANDOFF.md against actual repo state using the verify-then-update discipline — read the real git log, derived-doc state, and open worklist items FIRST, then update the one named handoff document to match, showing the assembled result before it's saved. Use at the end of a working session or before handing the project to another tool/person.

Generate the paged, Squarespace-ready HTML reader from the screenplay's text mirror via script_to_html_paged.py (Highland-calibrated pagination, renders [[ notes ]] as styled note boxes). Use to refresh the embeddable web reader after a script change.