How to plan meals from the recipes you've saved

Most people have recipes everywhere. There are links saved in browser bookmarks that have not been opened in months. There are screenshots buried in a camera roll. There is a handwritten card from a grandmother tucked inside a cookbook that never gets used because nobody can remember which cookbook. And every week, the same question arrives: what are we eating this week?

The problem is not a lack of recipes. It is the lack of a system that brings them together and makes them useful at the moment you need them. Fabric provides that system. It lets you collect recipes from any source, organise them in ways that suit how you cook, and then draw on the whole collection when you sit down to plan meals.

This guide walks through the full process, from saving your first recipe to building weekly meal plans with the help of Fabric's AI assistant.

Gathering recipes into one place

The first step is consolidation. Recipes need to leave the dozen apps, folders, and kitchen drawers where they currently live and arrive in a single, searchable library.

For anything you find online, the web clipper saves a full copy of the recipe page to Fabric. It captures the ingredients, method, and any photos, so you are not relying on the original website staying online or keeping the same URL. If you have been saving recipes from social media platforms, the clipper works there too. You can find a more detailed walkthrough in the guide on how to save recipes from Instagram and TikTok.

Physical recipes need a different approach. Fabric's quick capture lets you photograph a page from a cookbook or a magazine clipping. The text is extracted automatically, so the recipe becomes searchable alongside everything else in your collection. For handwritten recipes, the handwriting scanner handles the particular challenge of reading someone's personal script, whether that is your own hurried notes or a relative's cursive from decades ago. There is a separate guide on saving handwritten family recipes if you have a collection of those to work through.

You might also have recipe notes that exist only in your head: the tweak you made to a bolognese last Tuesday, or the way you adjust a curry depending on who is coming for dinner. Voice memos let you capture those modifications by speaking them aloud. Fabric transcribes and files them, connecting each note to the recipe it relates to.

Once everything is in Fabric, it is stored in your personal cloud, encrypted with AES-256 and CASA Tier 2 certification. Your recipes are yours, not training data for someone else's model.

Organising recipes so you can find them later

A pile of recipes in one place is better than a pile of recipes in twelve places, but it still needs structure. Fabric gives you two complementary tools for this: tags and folders.

Tags work well for attributes that cut across categories. You might tag recipes by protein (chicken, fish, lentils), by cuisine (Thai, Mexican, Italian), by dietary requirement (gluten-free, dairy-free, vegetarian), by season (summer salads, winter warmers), or by effort level (weeknight quick, weekend project). A single recipe can carry as many tags as it needs, so a quick vegetarian Thai curry sits comfortably under all four of those labels at once. The organisation features in Fabric make tagging fast, and once you have a consistent set of tags, filtering your collection becomes almost instant.

Folders serve a different purpose. They are good for groupings that change over time, like a folder for this week's meal plan, a folder for recipes you want to try next, or a folder for dishes that worked well at dinner parties. You can move recipes in and out of folders without losing their tags, so the underlying organisation stays intact while the weekly plan refreshes.

The combination matters. Tags describe what a recipe is. Folders describe what you are doing with it right now.

Using search to find the right recipe at the right moment

Tagging and folders give you structure, but sometimes you need to search sideways. You know you have a recipe that uses butternut squash and sage, but you cannot remember what it was called or where it came from. Fabric's semantic search handles this. Rather than matching exact words in a title, it understands the meaning behind your query. Searching for "something with root vegetables and pastry" will surface that butternut squash tart even if the word "tart" does not appear in your search.

This is particularly useful for meal planning because the questions you ask when planning are rarely "what is the name of that recipe?" They are more like "what can I make with the chicken thighs in the freezer?" or "what is a good side dish for grilled fish?" Semantic search lets you query your own collection the way you think about cooking.

Planning meals with the AI assistant

With your recipes collected, tagged, and searchable, you are ready to plan. This is where Fabric's AI assistant becomes particularly valuable, because it has access to your entire recipe library and can reason across it.

You can ask it things like "suggest five dinners for next week using what is in season in October" or "plan three nights of meals that use chicken, and make sure at least one is gluten-free." The assistant draws on the recipes you have saved, not a generic database, so its suggestions come from dishes you have already chosen to keep. It knows which recipes you have tagged as quick weeknight meals and which ones you have filed under weekend projects. It can also spot connections you might not have made yourself, suggesting that the leftover roast chicken from Sunday works well in the Vietnamese salad you saved last month.

Once you have a plan you are happy with, you can drop the selected recipes into a folder for the week. Some people create a new folder each week; others reuse a single "This Week" folder and swap recipes in and out.

How the memory engine ties everything together

Behind the scenes, the memory engine is building connections between your recipes and everything else in your Fabric library. If you saved an article about seasonal British produce alongside your recipes, the memory engine notices the relationship. If you clipped a blog post about fermenting vegetables and later saved a kimchi recipe, it links those together. Over time, your recipe collection becomes a web of related knowledge rather than a flat list.

This matters for meal planning because the assistant can draw on those connections. When it suggests meals, it is not just pattern-matching on ingredients. It understands that your interest in Middle Eastern cooking connects to the spice guide you bookmarked, which connects to the lamb shoulder recipe you photographed from a cookbook. The more you save, the richer the suggestions become.

Accessing your recipes from other tools

Fabric is not a walled garden. If you use ChatGPT or Claude for other parts of your life, you can give those tools scoped, time-limited access to your Fabric library through MCP or agent access. That means you could be in a conversation with ChatGPT about what to cook tonight and it can search your saved recipes directly, with your permission, without you needing to copy and paste anything. The agent access is designed so that external tools see only what you allow them to see, for only as long as you allow it.

This is useful if you already have workflows built around other AI tools and want your recipe collection to be available wherever you are thinking about food.

Frequently asked questions

Can I save recipes from any website?

Yes. The web clipper works on any public webpage. It saves the full content of the page, including images, so you have a permanent copy even if the original site changes or goes offline. For a broader look at saving content from across the web, see the guide on how to save everything you find online in one place.

How does Fabric handle recipes from videos?

When you save a recipe video from a platform like Instagram or TikTok, Fabric extracts the key information and stores it as searchable text alongside the video. You get both the visual reference and a version you can search and tag.

Can I share a meal plan with someone else?

You can share individual recipes or entire folders. If you build a weekly meal plan in a folder, you can share that folder with a partner or housemate so everyone can see what is on the menu.

Does the AI assistant suggest recipes I have not saved?

The assistant focuses on your own collection. It works with what you have saved and tagged, which means its suggestions reflect your taste and dietary needs rather than a generic recipe database.

How do I handle recipes with variations?

Voice memos are useful here. Save the base recipe as a clipped page or photograph, then record a voice note describing your preferred changes. Fabric links the note to the recipe, so when you pull it up later, the modification is right there.

Can I search by ingredient to use up what I have?

Yes. Semantic search lets you type something like "recipes using courgettes and feta" and Fabric will find matching recipes from your collection, even if those exact words do not appear in the recipe title.

Is my recipe collection private?

Everything in Fabric is stored in your personal cloud with AES-256 encryption and CASA Tier 2 certification. Your recipes are not used to train AI models or shared with third parties.

What if I have hundreds of recipes to import?

Start with the recipes you cook most often and build from there. The web clipper and quick capture make it fast to add new recipes as you encounter them, and over time your collection grows organically. There is no need to do everything at once.

Can I plan meals around dietary requirements?

Tags make this simple. If you tag recipes with dietary information (vegetarian, nut-free, low-FODMAP, or whatever applies), you can filter your collection by those tags when planning. The AI assistant also understands dietary constraints, so you can ask it to plan a week of dairy-free dinners and it will draw only on recipes that fit.

How does meal planning work across seasons?

Tag recipes by season or by the produce they feature. When autumn arrives, you can filter for recipes tagged with seasonal ingredients like squash, apples, or root vegetables. The AI assistant can also factor in seasonality if you ask it to suggest meals based on what is currently in season.

Related pages

• How to save recipes from Instagram and TikTok

• How to save handwritten family recipes

• How to save everything you find online in one place