mcp-parigp

MCP server exposing cypari2 (PARI/GP) number theory library

mcp-name: io.github.daedalus/mcp-parigp

Install

pip install mcp-parigp

Usage

As MCP Server

Run directly:

mcp-parigp

Or use with an MCP client by configuring in your settings:

{
  "mcpServers": {
    "mcp-parigp": {
      "command": "mcp-parigp"
    }
  }
}

In Python

from mcp_parigp import eval_expression, factor, isprime

# Evaluate PARI/GP expressions
result = eval_expression("factor(100)")
print(result)  # [[2, 2], [5, 2]]

# Factor integers
print(factor(100))  # [[2, 2], [5, 2]]

# Test primality
print(isprime(29))  # True

API

Number Theory

• factor(n) - Factor an integer
• isprime(n) - Test if n is prime
• gcd(a, b) - Greatest common divisor
• phi(n) - Euler's totient function
• sigma(n, k) - Sum of k-th power of divisors
• jacobi(a, n) - Jacobi symbol
• znorder(x, n) - Multiplicative order modulo n
• primes(n) - First n primes
• nextprime(n) - Next prime after n

Polynomials

• polroots(pol) - Find roots of polynomial
• polcyclo(n) - n-th cyclotomic polynomial
• deriv(pol) - Derivative of polynomial
• subst(pol, v, expr) - Substitute in polynomial

Number Fields

• nfinit(pol) - Initialize number field
• bnfinit(pol) - Initialize with Buchmann's algorithm
• idealadd(nf, a, b) - Add ideals
• idealmul(nf, a, b) - Multiply ideals

Elliptic Curves

• ellinit(eq) - Initialize elliptic curve
• ellap(E, p) - Trace of Frobenius at p
• elltors(E) - Torsion subgroup
• ellheight(E, P) - Canonical height

Matrices

• matid(n) - n×n identity matrix
• matdet(m) - Matrix determinant
• matinv(m) - Matrix inverse
• matrank(m) - Matrix rank

Elementary Functions

• sin, cos, tan, asin, acos, atan
• sinh, cosh, tanh, asinh, acosh, atanh
• exp, log, sqrt, abs
• pi(), euler(), I()

Development

git clone https://github.com/daedalus/mcp-parigp.git
cd mcp-parigp
pip install -e ".[test]"

# run tests
pytest

# format
ruff format src/ tests/

# lint
ruff check src/ tests/

# type check
mypy src/