olang-language-model

O Language Programming Model

A programming model specialized for the O Language programming language. Its primary purpose is to turn natural-language requests such as “write this in O Language” into runnable .ola code.

The model analyzes requirements, selects documented O Language syntax, and produces a complete .ola program. It can also explain, debug, and improve existing O Language code.

Capabilities

• Create new O Language programs
• Explain and improve .ola code
• Answer O Language syntax and command questions
• Generate examples for files, processes, networking, and system operations
• Use O Language documentation while generating code
• Convert natural-language requirements into .ola programs
• Produce .ola programs, .olm modules, and .olp package files
• Solve general programming and debugging requests in the language requested by the user

When a request explicitly names O Language or Olang, the response must use documented O Language syntax. For other programming requests, the requested programming language should be used.

Usage

Ask the model to solve a task in O Language. For code-generation requests, it should return a complete artifact instead of only an explanation.

Example request:

Write an O Language program that asks for a name and prints a greeting.

The canonical minimal example is:

output("Hello, World!")

Validate generated code with the O Language runtime before executing it.

Tool support

The model is designed to work through the agent layer with:

• search_docs: search the O Language documentation.
• explain_syntax: find usage for an O Language symbol or command.
• validate_olang: validate generated .ola code with the O Language runtime.
• analyze_olang: analyze likely syntax or runtime problems.
• run_olang: run documented run, exec, compile, or info operations.
• write_olang_artifact: write an explicitly requested .ola, .olm, or .olp file in the configured workspace.

Tool requests use JSON:

{"tool":"validate_olang","arguments":{"code":"output(\"Hello\")"}}

The model is not only a documentation question-answering bot; it is intended to translate programming requirements into O Language programs, analyze bugs, and propose corrected code.

RAG usage

The model works best with the live RAG index built from the O Language wiki and website documentation:

https://onix-project.com/O_Language

The GGUF file contains model weights only. RAG retrieval and tool execution are provided by the accompanying agent runtime.

Safety and limitations

Generated code must be treated as untrusted output. Run it in a sandbox with restricted network, filesystem, and process access.

When the O Language documentation or runtime changes, rebuild the RAG index and reassess generated answers.