SessionKnit (Python)

Part of LoopEngine — a runtime for defining and running AI agents through a transparent ReAct loop — handling durable, DAG-shaped session logs with crash-interruption detection. Works standalone too, no dependency on LoopEngine itself.

Ported line-for-line from the TypeScript implementation. See the root README for the pitch, the prior-art table, and the scope decisions — this file only covers what's specific to running the Python code.

Install

pip install -e ".[dev]"

Quickstart

PYTHONPATH=src python3 examples/quickstart.py

from sessionknit import FileStorage, SessionKnit, SessionEntry

sessionknit = SessionKnit(
    FileStorage("./sessions"),
    has_unresolved_tool_call=lambda m: m.has_tool_call,
    build_continuation=lambda m: Message(role="user", content="Continue from where you left off."),
)

await sessionknit.append(session_id, SessionEntry(id=id, parent_id=parent_id, message=message))

result = await sessionknit.resume(session_id)
# result.messages, result.resumed_after_interruption

Test

pytest

Status

Chain reconstruction, topology repair, the async write-behind queue, and interruption detection are real and tested. Published as sessionknit on PyPI.