Metadata-Version: 2.4
Name: ooty-codes
Version: 1.0.0
Summary: Terminal-accessible MCP and Agentic AI lab experiment codes
Author: Ooty Codes Contributors
License: MIT
Keywords: 8051,ARM7,microcontroller,agentic-ai,lab,education
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Environment :: Console
Classifier: Topic :: Education
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Dynamic: license-file
Dynamic: requires-python

# ooty-codes

A small terminal package containing 31 educational lab-code examples from the supplied Microcontroller and Programming (MCP) and Foundations of Agentic AI revision PDFs.

## Install

```bash
python -m pip install ooty-codes
```

## Commands

```bash
ooty show all exp
ooty display 1a
ooty display "mcp 1a"
ooty display "agentic 3"
ooty display GraphRAG
```

The catalogue has 21 MCP experiments (1a–10, including lettered experiments) and 10 Agentic AI experiments. If a number is ambiguous across groups, qualify it with `mcp` or `agentic`. Use `ooty --help` for usage.

## Scope and code notes

- MCP source examples include 8051 assembly, 8051 C, and ARM7/LPC21xx C. Displaying code does not assemble, compile, flash, or simulate it. Use a compatible 8051/ARM toolchain and hardware/simulator separately.
- Agentic AI examples are educational demonstrations. They do not call commercial LLM APIs or run official AgentBench/SWE-bench evaluations.
- Obvious line-wrap damage from PDF extraction has been cleaned where needed; program behavior is intended for teaching and may need adaptation to a specific board/compiler.
- Experiment 3a/3b preserve the comparison structure in the supplied record; verify carry/branch behavior in your assembler before using on hardware.
- MCP experiment 8's source demonstrates an ARM7 LED output on P0.7; despite its title mentioning LCD, the supplied PDF code does not implement an LCD driver.
- Agentic AI Experiment 1 is a concise CoT-style trace demonstration, not an LLM-generated private chain of thought. Experiment 10 is a local sample scorecard, not an official benchmark run.

## Development

```bash
python -m pip install -e .
python -m unittest discover -s tests
```

## Publish

See `PUBLISHING.md`. Check the package name on PyPI before publishing; publishing requires your own PyPI account and API token.
