MPCEnv

Defined in stdlib/sequre/mpc/env.codon

MPCEnv[TP] is the central runtime environment for all secure computation in Sequre. Every @sequre-annotated function receives it as its first argument (mpc).

Type parameter

Parameter	Description
TP	Unsigned integer type for MPC arithmetic (default: mpc_uint = UInt[192])

Fields

Field	Type	Description
pid	int	This party's ID (0 = trusted dealer, 1..N = compute parties)
local	bool	Whether running in local (forked) mode
stats	MPCStats	Operation counters and byte-transfer tracking
randomness	MPCRandomness	PRG streams and seed management
comms	MPCComms[TP]	Network I/O (send, receive, reveal, broadcast)
arithmetic	MPCArithmetic[TP]	Beaver-triple multiplication and partitioning
polynomial	MPCPolynomial[TP]	Lagrange interpolation, polynomial evaluation
boolean	MPCBoolean[TP]	Bit decomposition, boolean circuits
fp	MPCFP[TP]	Fixed-point truncation and normalization
mhe	MPCMHE[TP]	Multiparty homomorphic encryption (CKKS)

Initialization

Local mode (single machine)

from sequre import local, Sharetensor as Stensor

@local
def my_computation(mpc, x: int):
    s = Stensor.enc(mpc, x)
    # ... secure operations ...
    print(s.reveal(mpc))

my_computation(42)  # mpc is injected automatically

The @local decorator forks NUMBER_OF_PARTIES processes (default 3) on the same machine, each with its own MPCEnv.

Online mode (distributed)

from sequre import mpc, Sharetensor as Stensor

mpc = mpc()  # Reads party ID from sys.argv
s = Stensor.enc(mpc, 42)
print(s.reveal(mpc))

Run at each party:

SEQURE_CP_IPS=ip1,ip2,ip3 ./bin/sequre script.codon <pid>

MHE setup

By default, mpc() and @local call mpc.mhe.default_setup() which:

1. Generates per-party secret key shards
2. Runs collective public key generation (CKG)
3. Runs collective relinearization key generation (RKG)
4. Runs collective rotation key generation

Skip this with --skip-mhe-setup if only MPC is needed.

Sub-modules

mpc.arithmetic

Method	Description
multiply(a, b, modulus)	Beaver-triple secure multiplication
multiply_matmul(a, b, modulus)	Secure matrix multiplication
inner_prod(a, modulus)	Secure inner product

mpc.boolean

Method	Description
bit_decomposition(a, bitlen, ...)	Decompose a shared value into shared bits
demultiplexer(bits, modulus)	One-hot decode from shared bits

mpc.fp

Method	Description
trunc(a, modulus, k, m)	Fixed-point truncation

mpc.polynomial

Method	Description
lagrange_interp(x, y, modulus)	Lagrange interpolation on shared points
powers(x, power, modulus)	Compute powers of a shared value

mpc.mhe

Method	Description
default_setup()	Initialize CKKS params and generate collective keys
collective_init(params, prec)	Full collective initialization with custom params
enc_vector[T](values)	Encrypt a list into a list of Ciphertext or Plaintext

mpc.comms

Method	Description
send_as_jar(data, to_pid)	Serialize and send to a specific party
receive_as_jar(from_pid, T=type)	Receive and deserialize from a party
reveal(share, modulus)	Reconstruct a shared value
sync_parties()	Barrier synchronization
collect(value)	Gather a value from all parties

mpc.stats

Tracks all operations automatically:

Counter	Description
secure_mul_count	Number of secure multiplications
secure_matmul_count	Number of secure matrix multiplications
bytes_sent	Total bytes transmitted
rounds	Communication rounds
partitions_count	Beaver partitions performed
truncations_count	Fixed-point truncations
secure_bootstrap_count	HE bootstrapping operations
secure_mhe_mpc_switch_count	MPC ↔ MHE protocol switches (via E2S/S2E)

Context managers

AllowMPCSwitch

Enables automatic MPC ↔ MHE switching for the duration of the block. Inside this context, operations like Ciphertensor.matmul may automatically convert to Sharetensor (via the E2S protocol), run the operation using Beaver-triple MPC, and convert back (via S2E) when a cost estimator determines this is cheaper than a pure-HE path.

with mpc.allow_mpc_switch():
    result = ct_a @ ct_b  # matmul may use MPC path if cheaper

See the dedicated MPC ↔ MHE Switching page for details.

ModulusSwitch

Temporarily switch the working modulus (e.g., from field to ring).

StatsLog

Log operation statistics (multiplications, communication rounds, bytes sent, etc.) for a code block. Resets counters on entry and prints a summary on exit.

with mpc.stats_log("Training phase"):
    model.fit(mpc, X=X, y=y, step=0.1, epochs=10, momentum=0.9)

Write stats to a file instead of (or in addition to) stdout:

with mpc.stats_log("Inference", file_path="results/inference_stats.txt"):
    prediction = model.predict(mpc, X)

Parameter	Type	Default	Description
msg	str	""	Label printed alongside the stats summary
file_path	str	""	If non-empty, write stats to this file
mode	str	"a+"	File open mode
file_only	bool	False	If True, only write to file (suppress stdout)