Quilty Brings Screenplay Coverage Into Claude and ChatGPT -- and Scores the Rewrite Against the Last Draft

Directors and writers can now work their coverage notes inside the assistant they already use, then ask Quilty whether the new draft actually improved.

LOS ANGELES, Oct. 8, 2026 /PRNewswire/ -- Quilty (quilty.app) founders Emmy, Tony, and Olivier Award–winning producer Simon Horsman and technology entrepreneur Daniel Wood, announced today that the AI entertainment-intelligence platform for screenplay analysis is now available inside both Claude and ChatGPT. Writers can open a chat, connect their Quilty account, and work directly from coverage they already own — Quilty Score, story notes, character work, market and talent reads — without exporting a PDF or pasting a report into the window.

Alongside the integrations, Quilty highlights a capability aimed squarely at the rewrite: draft-over-draft comparison. A writer who runs a new draft through Quilty can ask the assistant to compare it to the previous analysis and get back the change in Quilty Score, pillar by pillar.

"Putting the analysis inside Claude and ChatGPT means a writer can take a note, try a new scene, and come back to Quilty to see whether the work actually moved the number. That is development, not a file sitting in downloads," said Horsman, CEO of Quilty.

Quilty Analysis is Built to read like a coverage reader, not a data dump. A full Quilty feature analysis runs 33 stages and produces a report with more than seventy section types across story craft, writer's craft, character and dialogue, market and finance, talent packaging, audience and culture, and production. The integrations were deliberately designed not to unload all of it at once.

Thirteen read-only tools let the assistant start where a professional reader starts — the Quilty Score and a bounded brief of strengths, issues, and priority actions — and then drill into a chapter, a single section, or a specific scene. Writers can run semantic search across their own screenplay text, search the analysis itself, pull the talent package or comparable-films research, or request a short-lived link to the full PDF.

"The full report is still there when you want the printed book. But Claude should not swallow seventy sections on the first ask. Score, main points, then drill down — that is how a coverage reader works, and it is how we built the server," said Wood, CTO of Quilty. 

Underneath the integrations is Quilty Score V5, a 0–100 composite built from four weighted pillars — Story & Craft, Commercial Viability, Cultural Resonance, and Production Reality — broken into fifteen sub-dimensions. Scores are produced by an ensemble of models, reported with a confidence interval and a model-agreement figure rather than a single confident number.

Scoring runs blind by default: identifying material is stripped from the screenplay before it reaches the scoring models, so a well-known title is judged on the pages rather than on its reputation. Quilty Scores are diagnostic assessments of the submitted material, not predictions of box office.

That engine is fed by a deliberately multi-vendor stack. Quilty routes each stage to the model best suited to it, drawing on Anthropic's Claude, OpenAI's GPT, xAI's Grok, Google's Gemini, Perplexity, and DeepSeek — creative and craft judgment to one family, cultural and financial reasoning to another, live market research to a search-native model. No single vendor's blind spot becomes Quilty's blind spot.

More information here:
https://drive.google.com/drive/folders/13oBHVh16vczU-Sy7yhA6JMqo6fqKwgun?usp=sharing 

• How it Works. Step-by-step.
• Insight from a Quilty user, writer/director/producer Nick Lyon.
• Press release WORD and PDF versions

About Quilty 

Integration documentation: quilty.app/docs/claude-connector
Claude Desktop and Claude Code connect with a personal token from Settings.
Connectors on quilty.app. Same server, same scope: https://quilty.app/api/mcp.

Contact:  David Roberson
Roberson Public Relations
[email protected]
310-925-8786

SOURCE Quilty