icebox-gov 0.2.0

Runtime governance for autonomous offensive security — the single seam every human operator, REST client, and LLM agent must pass through before anything touches a target.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

"""Python SDK for the ICEBOX Governance Kernel.

``pip install icebox-sdk`` gives you :class:`Workspace` for high-level
orchestration, and :class:`IceboxClient` for talking to the REST API.
"""

from ._sdk import Governance, IceboxClient, IceboxError

class Workspace:
    """High-level abstraction for ICEBOX orchestration.
    
    Wraps the ICEBOX engine so AI agents don't have to deal with raw
    C-pointers or HTTP endpoints directly.
    """

    def __init__(self, target: str, url: str = "http://127.0.0.1:8443"):
        """Initialize the workspace with a target path and ICEBOX daemon URL."""
        self.target = target
        self.client = IceboxClient(url)
        self.client.accept_charter(self.target)
        self.client.add_scope(self.target)
    
    def execute(self, module: str, sandbox: bool = False, approved: bool = True, options: dict | None = None) -> dict:
        """Executes a module against the workspace target."""
        return self.client.run_module(module, self.target, sandbox=sandbox, approved=approved, options=options)

    def audit(self, n: int = 20) -> list[dict]:
        """Retrieves the JSON audit trail for the workspace."""
        return self.client.audit(n)

__all__ = ["Governance", "IceboxClient", "IceboxError", "Workspace"]
__version__ = "0.2.0"