# Grant Context Files Structure This folder contains PDFs, HTML files, and other reference documents for Innovate UK grants. ## Folder Structure ``` context/ ├── general/ │ ├── innovate-uk-info.html │ ├── application-guidelines.html │ └── [other general IUK info] │ └── grants/ ├── competition-1389/ │ ├── guidance.pdf │ ├── terms.pdf │ ├── scope.html │ └── [supporting documents] ├── competition-2006/ │ └── [supporting documents] ├── competition-2010/ │ └── [supporting documents] └── [... 28 grant folders total ...] ``` ## How to Use ### Option 1: Upload PDFs 1. Save grant PDFs from Innovate UK website 2. Place in the appropriate `grants/[competition-id]/` folder 3. Claude Code can read PDFs directly ### Option 2: Save Supporting Pages as HTML 1. Right-click on supporting pages → Save As → Complete HTML 2. Save to `grants/[competition-id]/` 3. Claude Code can extract info from HTML ### Option 3: Let Claude Code Scrape 1. Provide URLs of supporting pages 2. Claude Code will extract and organize them ## Current Folders (28 Total) All grant folders have been created and are ready for documents: - competition-1389 - competition-2006 - competition-2010 - competition-2058 - competition-2177 - competition-2185 - competition-2186 - competition-2265 - competition-2275 - competition-2276 - competition-2279 - competition-2284 - competition-2288 - competition-2293 - competition-2295 - competition-2298 - competition-2299 - competition-2305 - competition-2307 - competition-2313 - competition-2314 - competition-2315 - competition-2316 - competition-2317 - competition-2318 - competition-2322 - competition-2327 - competition-2328 ## Next Steps 1. **Gather PDFs**: Download or save grant supporting documents from Innovate UK 2. **Organize**: Place each document in its corresponding `grants/[id]/` folder 3. **Reference**: When Claude Code needs grant details, it can read these files 4. **Scale**: As you add more grants, create new folders following the same pattern ## Notes - Claude Code can read: PDF, HTML, TXT, MD, and other text formats - No need to manually convert or transcribe - just upload the files - The model will intelligently extract relevant information - HTML files are recommended for web pages (smaller, searchable) - PDFs work great for official documents (guidance, terms, conditions)