Skip to main content

Local enforcement SDK and simulation environment for AI governance.

Project description

Waveframe Guard

Stop unsafe AI and automated actions before they execute.

Waveframe Guard enforces governance at the execution boundary. If an action violates policy, it never runs.

Example

from waveframe_guard import install_guard, guard
from compiler.compile_policy import compile_policy

policy = {
    "contract_id": "finance-core",
    "contract_version": "0.3.0",
    "authority": {
        "required_roles": ["manager"]
    }
}

compiled = compile_policy(policy)

install_guard(
    actor={"id": "user-1", "type": "human", "role": "intern"},
    contract=compiled
)

@guard
def transfer(amount):
    print(f"Transferred ${amount}")

transfer(100)
Execution blocked: required role not satisfied: manager

What Waveframe Guard Does

  • Intercepts function execution
  • Evaluates governance rules before execution
  • Blocks invalid actions deterministically
  • Continues enforcement even if Cloud is unavailable

Local vs Cloud

Mode Behavior
Local Fast, local enforcement
Cloud Policy sync, audit, and attestation

Guard enforces locally. Cloud provides authority, audit, and verification.

Fail Modes

Mode Behavior
cache (default) Use cached policy if Cloud is unavailable
closed Block if Cloud is unavailable and no cached policy exists
open Allow execution if policy is unavailable and mark the decision unverified

Install

pip install waveframe-guard cricore-contract-compiler cricore-proposal-normalizer

Live Demo

python examples/live_enforcement_demo.py

The demo shows:

  • an intern blocked by policy
  • a manager allowed by policy
  • cached local enforcement during a simulated Cloud outage

Why This Exists

Most AI systems can suggest, warn, or log.

Waveframe Guard is the layer that can stop execution.

Architecture Note

The Waveframe Guard SDK operates independently and does not require Cloud components to enforce governance locally.

The cloud directory contains experimental Cloud control plane components and is not required for SDK operation.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

waveframe_guard-0.3.0.tar.gz (8.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

waveframe_guard-0.3.0-py3-none-any.whl (7.7 kB view details)

Uploaded Python 3

File details

Details for the file waveframe_guard-0.3.0.tar.gz.

File metadata

  • Download URL: waveframe_guard-0.3.0.tar.gz
  • Upload date:
  • Size: 8.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.4

File hashes

Hashes for waveframe_guard-0.3.0.tar.gz
Algorithm Hash digest
SHA256 a2058d9b058263234ce7820721b4a03ef15400bc738d42a7772e90af0b913a58
MD5 55921710fbe0e70f0e58ea7307ef62bd
BLAKE2b-256 85493e99d993842089fbbb413cae03ef62ca3addb9a8eb1c1dcefdbacf9dae01

See more details on using hashes here.

File details

Details for the file waveframe_guard-0.3.0-py3-none-any.whl.

File metadata

File hashes

Hashes for waveframe_guard-0.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 ec8df42b06fb16ed2bbf54da348dd94f052119279635145e8068f29e0a2aeadd
MD5 331e5400a0d6a2304c5590211b2f3194
BLAKE2b-256 57be48b18527709276b91bbcbc054fa7872e5baac1fb144276047070f1af7dec

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page