# NotCo MCP

> NotCo connects your AI agent directly to Giuseppe, a proprietary Food Tech platform. Use it to run molecular analysis on ingredients, match flavor profiles, predict nutritional values, and generate full plant-based formulations for B2B R&D. It's built for product formulators needing to test recipes virtually before the lab.

## Overview
- **Category:** artificial-intelligence
- **Price:** Free
- **Tags:** food-tech, plant-based, molecular-analysis, r-and-d, recipe-generation

## Description

Yo, check this out: NotCo connects your AI agent straight into **Giuseppe**, which is their proprietary Food Tech platform. You're running advanced food science simulations here—it’s how you accelerate B2B research and development using nothing but natural language commands.

This thing handles the whole product lifecycle virtually, from screening raw materials to predicting final costs. Forget physical testing; you test recipes right in your agent. It's built for formulators who need to blueprint a recipe before they even touch a lab bench.

When you wanna build something new, you’ll use `create_formulation` to ask Giuseppe to generate an entire plant-based formula blueprint based on specific constraints and the target product. You can then pull up details on any existing concept using `get_formulation`, or if you need a list of what's out there, run `list_formulations` to see all current AI formulations across categories like dairy or meat.

Before you start building, you gotta know your parts. Use `list_ingredients` to search the molecular database for every available ingredient, and then use `get_ingredient` on a specific entry to get its full profile—that includes its molecular structure, sensory data, and functional details. You can also run `list_suppliers` to see which approved vendors you're working with.

Need benchmarks? Use `list_nutritional_profiles` to pull up target industry standards so you know what your final recipe needs to hit. Similarly, if you want a baseline for tracking progress, use `list_projects` or `create_project` to start and manage a structured R&D project.

Let's talk about the science stuff. If you wanna see how good something is nutritionally, run your formula through `analyze_nutrition`. It predicts the final output—calories, fat, carbs, and more—per 100g. Wanna know what it’ll cost to make? Use `estimate_cost` to predict the theoretical mass-production price based on global commodity rates for that mix. You can also check out all the available sensory standards by running `list_sensory_profiles`. 

If you're trying to nail a specific flavor, don't guess. Run `search_flavor_matches` to find plant combinations that actually replicate a target volatile molecule or aroma. To simulate what it’ll taste like for actual consumers, use `run_sensory_test`; this runs an AI simulation of the sensory experience on your generated formula. You can also run through `list_formulations` if you want to compare existing options against these new virtual results.

It's a full molecular toolkit: You start by defining what you need, then you pull data on ingredients and standards using `get_ingredient`, `list_ingredients`, and `list_nutritional_profiles`. Next, you ask the system to build it with `create_formulation` or check an old one with `get_formulation`. After that, you test everything: Run through `analyze_nutrition` for metrics, calculate cost with `estimate_cost`, match flavors using `search_flavor_matches`, and predict the taste profile with `run_sensory_test`.

## Tools

### analyze_nutrition
Predicts the nutritional output (calories, fat, carbs) of a specific product formulation.

### create_formulation
Requests and generates an entirely new AI-driven formula based on user prompts and constraints.

### create_project
Initiates a structured R&D project within the system to track related formulations and tests.

### estimate_cost
Predicts the theoretical mass-production cost of an existing or generated formula based on commodity prices.

### get_formulation
Retrieves all detailed information about a previously created or specified AI formulation ID.

### get_ingredient
Gets the complete molecular, sensory, and functional profile for one specific ingredient.

### list_formulations
Lists existing plant-based AI formulations across different product categories (dairy, meat, etc.).

### list_ingredients
Searches and retrieves a list of available ingredients from the molecular database.

### list_nutritional_profiles
Retrieves target nutritional benchmarks so you can compare formulas against industry standards.

### list_projects
Lists all active or completed R&D projects to track overall progress.

### list_sensory_profiles
Retrieves standard, pre-defined sensory profiles for comparison against generated formulas.

### list_suppliers
Lists approved and verified ingredient suppliers for sourcing data.

### run_sensory_test
Runs an AI simulation of a specific sensory test on a given formulation to predict consumer experience.

### search_flavor_matches
Finds combinations of plant ingredients that mimic or replicate a target flavor profile.

## Prompt Examples

**Prompt:** 
```
Ask Giuseppe to create a plant-based alternative for condensed milk, with the constraint 'no palm oil'.
```

**Response:** 
```
Giuseppe generated Formulation ID #CM-882. The top ingredient matches include: Coconut Oil (35%), Faba Bean Protein isolate (18%), Cabbage Extract (for molecular lactose emulation - 2%), and Beet fiber. The theoretical match score on viscosity and flavor is 92.4%.
```

**Prompt:** 
```
Run a nutritional analysis on formulation ID #CM-882.
```

**Response:** 
```
Predictive nutritional analysis for Formulation #CM-882 (per 100g): Calories: 325kcal, Total Fat: 11g (Saturated 8g, Trans 0g), Total Carbohydrate: 55g (Sugars 54g), Protein: 3g. Compared to standard dairy condensed milk, this formulation achieves a 15% reduction in saturated fats while maintaining identical sugar curves.
```

**Prompt:** 
```
Estimate the mass-production cost for this formulation.
```

**Response:** 
```
Based on current global commodity rates for the specified plant extracts (Coconut Oil, Faba Bean, etc.), the estimated cost for Formulation #CM-882 is $1.15 USD per kg. This represents a 12% cost advantage compared to the current dairy milk spot pricing.
```

## Capabilities

### Generate New Formulas
Ask Giuseppe to create a full, detailed plant-based recipe blueprint by specifying the target product and constraints.

### Analyze Nutrition Data
Run a formulation through predictive algorithms to get estimated nutritional values (calories, fat, protein, etc.) per 100g.

### Determine Ingredient Profiles
Retrieve the complete molecular and sensory data for any known plant extract or protein.

### Predict Manufacturing Cost
Calculate the theoretical mass-production cost of a formulated recipe based on global commodity pricing.

### Match Flavor Molecules
Use computational models to find plant combinations that replicate specific volatile molecules, aromas, or mouthfeels.

## Use Cases

### Creating a dairy alternative with specific fat constraints.
A product developer needs condensed milk without palm oil. They ask their agent to `create_formulation` for the target profile. The system uses over 300,000 plant records and returns an initial formula ID. Next, they run `analyze_nutrition` on that ID to confirm the fat reduction percentage while maintaining required sugar curves.

### Verifying the source of a novel ingredient.
A chemist finds a promising new extract but needs to know its molecular makeup. They use `get_ingredient` with the name, which immediately returns the full profile and confirms if it's available from an approved supplier listed by `list_suppliers`.

### Simulating the consumer experience for a new meat substitute.
Before running expensive sensory panels, the team uses `run_sensory_test`. They input the preliminary formula ID and get an AI simulation score on texture and aroma. This saves time and directs immediate adjustments to the initial mix.

### Evaluating market feasibility for a new line.
A strategic partner generates five different formulas across various product types. They use `estimate_cost` on all of them. The agent compares these costs against existing global commodity rates, instantly flagging which formulations are most profitable.

## Benefits

- Stop guessing about shelf stability. Use `analyze_nutrition` to predict final nutritional values, ensuring your plant-based formula hits specific health targets before you waste time in the lab.
- Slash R&D cycles by predicting costs upfront. Running `estimate_cost` gives you a theoretical mass-production price point, letting strategic partners evaluate market viability instantly.
- Go beyond simple ingredient lists. With `get_ingredient`, your agent pulls the full molecular profile—including sensory and functional data—so you know exactly what compounds are in play.
- Don't rely on trial and error for flavor. Use `search_flavor_matches` to find plant combinations that replicate a specific aroma or mouthfeel, speeding up the concept phase dramatically.
- Keep everything organized by using `list_projects`. This tool allows you to track all related formulations, costs, and tests under one master R&D umbrella.

## How It Works

The bottom line is that you don't need to switch between spreadsheets and databases; your agent handles the entire flow from concept query to finalized data report.

1. Subscribe to the NotCo server and input your required API Key.
2. Call a foundational tool (e.g., `list_ingredients`) using your AI client to gather raw data or define project scope.
3. Pass the resulting ingredient lists, profiles, or parameters into an analysis tool (like `create_formulation` or `analyze_nutrition`) to get the final blueprint.

## Frequently Asked Questions

**How does analyze_nutrition work with create_formulation?**
You must run `analyze_nutrition` *after* you use `create_formulation`. The formula generation provides the recipe ID, and this tool uses that ID to predict the final nutritional values based on current data.

**Can I find out the cost of a formula using estimate_cost?**
Yes. You call `estimate_cost` and provide it with a formulation ID or ingredient list. It uses global commodity pricing to predict the theoretical mass-production cost.

**What is the difference between list_ingredients and get_ingredient?**
`list_ingredients` gives you a searchable roster of available ingredients in the database. `get_ingredient` pulls the full, detailed molecular profile for one specific ingredient ID.

**Do I need to create_project before running any formulas?**
It's best practice. Calling `create_project` first establishes a single container ID. You then run all subsequent tools (like `run_sensory_test`) against that project ID for clean tracking.

**How does search_flavor_matches help my product?**
This tool finds plant combinations that mimic a target flavor or aroma. It's crucial early in R&D because it gives you actionable ingredient suggestions instead of generic ideas.

**When using list_ingredients, how do I handle rate limits or bulk requests?**
The server enforces standard rate limits per API key. For high-volume data pulls, you must implement exponential backoff logic in your client. We recommend caching results for a defined period to manage load.

**If I use analyze_nutrition with a novel mix of ingredients, what does the tool do if some components are outside the database?**
The tool flags external inputs and calculates only the metrics for known compounds. It returns an estimated value for the unknown component, along with a confidence score indicating data uncertainty.

**Does list_suppliers allow me to filter by geographical region or specific compliance standards?**
Yes, you pass geo-filters and compliance codes in the request body. This lets your agent narrow down suppliers immediately—for example, restricting results only to EU-certified sources.

**Can I use this to generate a formula for a new plant-based meat?**
Yes! Use the `create_formulation` tool to instruct Giuseppe to model a specific animal product (e.g., 'Target: Pulled Pork'). You can pass constraints like 'no soy' or 'must contain pea protein'. Giuseppe will return a mathematically generated formula combining plant extracts that mimic the target's molecular profile.

**How does Giuseppe match specific flavors?**
The `search_flavor_matches` tool analyzes NotCo's proprietary database. Instead of searching for 'beef flavor', Giuseppe searches for the specific volatile molecular compounds that create the beef flavor, and then finds combinations of seemingly unrelated plants (like pineapple and cabbage) that, when combined mathematically, replicate that exact molecular behavior.

**Can I predict the cost of a new formulation before making it?**
Absolutely. Once Giuseppe generates a formulation, you can pass its ID to the `estimate_cost` tool. The API cross-references the required plant ingredients with global B2B commodity pricing databases to give you an estimated per-kilogram cost for mass production.