Kissrule

This is a lightweight rule-based system for game AI micro-decisions, written in C++20. It lets you register C++ object hierarchies as named facts, load rule sets from JSON, and evaluate them at runtime with lazy fact fetching and caching.

Features

Rules defined as flexible lists of criteria evaluating facts, fully serializable to JSON.
Rich comparison operators for criteria (equal, range, outside range, etc.).
Flexible fact source definitions.
Enum support with typed enum values in rules.
Multiple rule ordering policies (by priority, criterion count, or combined).
Multiple response execution policies with runtime filtering.
Performant runtime evaluation.

Coding Convention

We're loosely following the LLVM coding standards: https://llvm.org/docs/CodingStandards.html, but use snake_case for everything: types, variables and functions. Private member variables end with an underscore: member_variable_. Template parameters are framed with two underscores: _type_.

Known Limitations

Rule compilation is heavyweight, probably has room for improvement on the performance side.
64-bit value types only allow equality comparisons, ordinal comparisons are not supported (because every fact value is converted to double-precision floating point at rules compilation).
For now we're only joining criteria with AND operator. We may some day consider allowing OR operators in edited/serialized rules, and flattening them at compilation time.

To-do List

Error handling / data validation ++

Future Plans

Standalone rules editor (Qt6 DLLs + vcpkg, or maybe C#).
Godot integration (GDExtension, should be fine with C++20, even if the engine itself is C++17 -- at least according to Claude).
Unreal integration (a plugin, most likely).
CMake FetchContent to avoid storing third party libs in the repo, see for example https://github.com/eyalamirmusic/ScopedMemoryAllocations.