CoCo: Combined Deduction and Abduction Reasoner

CoCo (Combined deduCtiOn and abduCtiOn) is a dual-process cognitive architecture written in Rust. It combines a CLIPS-based rule engine with timeline planning to support both deduction and abduction in dynamic environments.

Highlights

• Hybrid reasoning: deduction + abduction
• Rust core for performance and safety
• CLIPS integration for rules and pattern matching
• Simple web interface for interaction and visualization

Quick start

Run the server locally:

cargo run --features="server"

By default the server listens on http://0.0.0.0:3000 and serves the web UI.

Installation

Prerequisites

CoCo links to the CLIPS v6.4.2 C library. Install CLIPS as follows:

1. Download and unpack:

wget -O clips_core_source_642.zip https://sourceforge.net/projects/clipsrules/files/CLIPS/6.4.2/clips_core_source_642.zip/download
unzip clips_core_source_642.zip

2. Build the core library:

cd clips_core_source_642/core
make release

3. Either point CLIPS_SOURCE_DIR at the extracted source, or install the static library system-wide:

export CLIPS_SOURCE_DIR=$(pwd)
# or
sudo cp libclips.a /usr/local/lib/
sudo ldconfig

Examples: loading Classes, Objects and Rules

Below are minimal examples that show JSON payloads (useful for APIs or config files).

1. Class example (JSON)

{
   "name": "Person",
   "parents": ["Agent"],
   "static_properties": {
      "age": { "type": "int", "default": 30, "min": 0, "max": 150 },
      "name": { "type": "string", "default": "Unknown" }
   },
   "dynamic_properties": {
      "mood": { "type": "symbol", "allowed_values": ["happy", "neutral", "sad"], "default": "neutral" }
   }
}

2. Object example (JSON)

{
   "classes": ["Person"],
   "properties": {
      "age": 28,
      "name": "Alice"
   }
}

3. Rule example (JSON)

{
   "name": "greet",
   "content": "(defrule greet (Person (name ?n)) => (printout t \\\"Hello \\\" ?n \\\"!\\\" crlf))"
}

Contributing

Contributions are welcome! Please open issues for bugs or feature requests, and submit pull requests for improvements.

License

CoCo is licensed under the MIT License. See LICENSE for details.