I’m building a small application around Claude Sonnet 4.6 (Thinking) and need a developer who can wire the model into a clean workflow that lets users generate brand-new text documents—or an entire folder of them—up to a 100 MB cap. Here’s the core flow I have in mind: a user enters a prompt, the app calls the Claude Sonnet API, and the returned content is saved to disk either as a single .txt/.md file or as a structured folder that can be zipped for download. No image handling is required; the emphasis is strictly on text generation from scratch. Key points to cover • Claude Sonnet 4.6 API integration, including secure key handling and streaming where supported • File-size guardrail so total output never exceeds 100 MB • Automatic packaging of multi-file results into a zip when the output spans a folder • Simple, intuitive UI (web or desktop—whichever stack you’re fastest with) plus a minimal backend or local wrapper service • Clear, well-commented code and a README that lets me re-deploy the project and swap API keys without hassle Acceptance criteria 1. Prompt → click → downloaded text file/folder in under one minute for a 50 MB test case. 2. Fails gracefully with a helpful message if the generated output would breach the 100 MB limit. 3. Runs on macOS or Windows without extra paid dependencies beyond the Claude API. If you’ve worked with Anthropic, OpenAI, or similar LLM endpoints and are comfortable building lightweight file-management utilities, this should be a straightforward engagement. Let me know your preferred tech stack and any questions you have about the workflow, and we can get started right away.