USolver
A Model Context Protocol server that exposes tools for solving combinatorial, convex, integer programming, and non-linear optimization problems. Exposes interfaces to the following solvers:
highs - Linear and mixed-integer programming solverortools - Combinatorial optimization solvercvxpy - Convex optimization solverz3 - SMT solver over booleans, integers, reals, and strings
To install run the install.py script. This will install the MPC server for Claude Desktop and/or Cursor.
Then open Claude or Cursor and you should see the MCP tool usolver available in the tool list.
Examples
To run the individual solver examples you can invoke the individual example modules. Each module contains a docstring which can be used to prompt the language model to solve the problem.
- Chemical Engineering - Pipeline design optimization using Z3 SMT solver for fluid transport systems with flow continuity, pressure drop, and economic constraints
- Chained Solvers - Multi-stage restaurant optimization combining OR-Tools for table layout and CVXPY for staff scheduling
- Job Shop Scheduling - Complex scheduling problem using OR-Tools to minimize makespan while respecting operation precedence and machine capacity
- Logistics - Transportation network optimization using HiGHS to minimize shipping costs in multi-stage supply chains
- Nurse Scheduling - Hospital staff scheduling using OR-Tools to assign nurses to shifts with fairness and availability constraints
- Portfolio Theory - Modern portfolio optimization using CVXPY to maximize returns while constraining risk across asset classes
- Production Planning - Manufacturing optimization using HiGHS to maximize profit subject to machine, labor, and material constraints
- Resource Allocation - Project portfolio selection using HiGHS mixed-integer programming to maximize value within budget and resource limits
- Network Flow - Shortest path optimization using HiGHS linear programming to find minimum-cost routes through directed graphs with flow conservation constraints
- Coin Problem - Classic logic puzzle using Z3 to find which 6 US coins total $1.15 but cannot make change for various denominations
- Cryptarithmetic - Solve cryptarithmetic puzzles like SEND + MORE = MONEY using Z3 constraint programming
- Knapsack Problem - Classic 0/1 knapsack optimization using OR-Tools to maximize value within weight constraints
- Multilinear Optimization - Linear programming with mixed constraints using Z3 to minimize objective functions subject to linear inequalities
- N-Queens - Place N queens on an N×N chessboard using OR-Tools constraint programming with no attacking positions
- Sparse Solver - Large-scale optimization demonstrating sparse matrix formats for memory-efficient resource allocation across facilities and time periods
Logic Puzzle
You and a friend pass by a standard coin operated vending machine and you decide to get a candy bar.
The price is US $0.95, but after checking your pockets you only have a dollar (US $1) and the machine
only takes coins. You turn to your friend and have this conversation:
You: Hey, do you have change for a dollar?
Friend: Let's see. I have 6 US coins but, although they add up to a US $1.15, I can't break a dollar.
You: Huh? Can you make change for half a dollar?
Friend: No.
You: How about a quarter?
Friend: Nope, and before you ask I cant make change for a dime or nickel either.
You: Really? and these six coins are all US government coins currently in production?
Friend: Yes.
You: Well can you just put your coins into the vending machine and buy me a candy bar, and I'll pay you back?
Friend: Sorry, I would like to but I can't with the coins I have.
What coins are your friend holding?
This can be fed into usolver and it will generate a constraint system:
Loading... is the collection of the six unknown coin values, Loading... through Loading..., each of which must be a positive whole number representing cents.
Loading...Loading... is the collection of every possible way you could choose two or more of your six coins.
Loading...Exclude the 50 cent coin from being used in the vending machine.
Loading...Constraint 0: The sum of the values of all six coins is 115 cents.
Loading...Constraint 1: Cannot make change for a dollar.
Loading...Constraint 2: Cannot make change for half a dollar.
Loading...Constraint 3: Cannot make change for a quarter.
Loading...Constraint 4: Cannot make change for a dime.
Loading...Constraint 5: Cannot make change for a nickel
Loading...Constraint 6: Cannot buy the candy bar for 95 cents if half dollar is excluded.
Loading...If you feed this to solver it will synthesize the above constraint system, solve it with Z3, and return the solution.
Your friend has: 1 half dollar, 1 quarter, and 4 dimes
This totals 50¢ + 25¢ + 40¢ = 115¢ = $1.15 ✓
This is exactly 6 coins ✓
Modern Portfolio Theory
A finance example:
Objective: Maximize expected portfolio return
Constraints:
Bonds allocation cannot exceed 40%
Stocks allocation cannot exceed 60%
Real Estate allocation cannot exceed 30%
Commodities allocation cannot exceed 20%
All allocations must be non-negative
Total allocation must equal exactly 100%
Total weighted portfolio risk cannot exceed 10%
Given Data:
Expected returns: Bonds 8%, Stocks 12%, Real Estate 10%, Commodities 15%
Risk factors: Bonds 2%, Stocks 15%, Real Estate 8%, Commodities 20%
This is compiled by the langauge model down into a convex optimization problem that can be cvxopt.
Loading...Where:
- Loading... = Bonds allocation
- Loading... = Stocks allocation
- Loading... = Real Estate allocation
- Loading... = Commodities allocation
The answer is then:
Bonds: 30.0%
Stocks: 20.0%
Real Estate: 30.0% (at maximum allowed)
Commodities: 20.0% (at maximum allowed)
Maximum Expected Return: 10.8% annually
Z3
A chemical engineering example:
Use usolver to design a water transport pipeline with the following requirements:
* Volumetric flow rate: 0.05 m³/s
* Pipe length: 100 m
* Water density: 1000 kg/m³
* Maximum allowable pressure drop: 50 kPa
* Flow continuity: Q = π(D/2)² × v
* Pressure drop: ΔP = f(L/D)(ρv²/2), where f ≈ 0.02 for turbulent flow
* Practical limits: 0.05 ≤ D ≤ 0.5 m, 0.5 ≤ v ≤ 8 m/s
* Pressure constraint: ΔP ≤ 50,000 Pa
* Find: optimal pipe diameter and flow velocity
Linear Programming
A multilinear optimization example:
Use usolver to solve the following linear programming problem:
Minimize: 12x + 20y
Subject to: 6x + 8y ≥ 100
7x + 12y ≥ 120
x ≥ 0
y ∈ [0, 3]
This is compiled by the language model down into a constraint satisfaction problem that can be solved with Z3.
Loading...Where:
- Loading... = First decision variable (continuous, non-negative)
- Loading... = Second decision variable (continuous, bounded)
The optimal solution is:
x = 15.0
y = 1.25
Objective value = 205.0
CVXPY
A simple convex optimization problem minimizing the 2-norm of a linear system:
Use usolver to solve the following convex optimization problem:
Minimize: ||Ax - b||₂²
Subject to: 0 ≤ x ≤ 1
where
A = [1.0, -0.5; 0.5, 2.0; 0.0, 1.0]
b = [2.0, 1.0, -1.0]
A classic worker shift scheduling problem:
Use usolver to solve a nurse scheduling problem with the following requirements:
* Schedule 4 nurses (Alice, Bob, Charlie, Diana) across 3 shifts over (Monday, Tuesday, Wednesday)
* Shifts: Morning (7AM-3PM), Evening (3PM-11PM), Night (11PM-7AM)
* Each shift must be assigned to exactly one nurse each day
* Each nurse works at most one shift per day
* Distribute shifts evenly (2-3 shifts per nurse over the period)
* Charlie can't work on Tuesday.
Chained Examples
A chained example that uses both OR-Tools to optimize for table layout and CVXPY to optimize for staff scheduling.
Use usolver to optimize a restaurant's layout and staffing with the following requirements in two parts. Use combinatorial optimization to optimize for table layout and convex optimization to optimize for staff scheduling.
* Part 1: Optimize table layout
- Mix of 2-seater, 4-seater, and 6-seater tables
- Maximum floor space: 150 m²
- Space requirements: 4m² (2-seater), 6m² (4-seater), 9m² (6-seater)
- Maximum 20 tables total
- Minimum mix: 2× 2-seaters, 3× 4-seaters, 1× 6-seater
- Objective: Maximize total seating capacity
* Part 2: Optimize staff scheduling using Part 1's capacity
- 12-hour operating day
- Each staff member can handle 20 seats
- Minimum 2 staff per hour
- Maximum staff change between hours: 2 people
- Variable demand: 40%-100% of capacity
- Objective: Minimize labor cost ($25/hour per staff)
Docker Usage
Can also run the MCP server directly from the GitHub Container Registry.
docker run -p 8081:8081 ghcr.io/sdiehl/usolver:latest
Then add the following to your client:
{
"mcpServers": {
"sympy-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"-p",
"8081:8081",
"--rm",
"ghcr.io/sdiehl/usolver:latest"
]
}
}
}
License
Released under the Apache License 2.0. See the LICENSE file for details.