Automate Everything

Practical automation scripts for developers who want local AI workflows, safer file operations, and repeatable command-line utilities they can inspect before running.

This repo is intentionally script-first: small tools, clear commands, safe defaults, and lightweight tests. The strongest current areas are Ollama setup, local model tooling, prompt utilities, semantic search, and file management helpers.

---

Table of Contents

• Why Use This
• What's Inside
• Quick Start
• Local AI Recipes
• Quality Gates
• Roadmap
• Contributing
• License

---

Why Use This

• Readable by default: every script is meant to be opened, understood, and adapted.
• Safer automation: file-moving and setup helpers prefer explicit inputs, dry runs, and clear failures.
• Local AI friendly: Ollama, llama.cpp, Hugging Face model downloads, Modelfiles, endpoint checks, and benchmarking are covered.
• Contributor ready: tests, linting, shell checks, issue templates, PR checklist, security notes, and examples are already in place.

---

What's Inside

Area	Scripts	Docs
Python file management	create_folder_structure.py, move_files.py	Python/README.md
Python AI utilities	token_counter.py, prompt_builder.py, batch_inference.py, embedding_search.py	Python/ai_utils/README.md
Python local AI	benchmark_ollama.py, download_hf_model.py, generate_modelfile.py, ollama_model_report.py	Python/local_ai/README.md
Bash Ollama setup	install_ollama.sh, pull_models.sh, run_inference.sh	Bash/README.md
Bash local AI	check_gpu.sh, install_llama_cpp.sh, setup_python_ai_env.sh	Bash/local_ai/README.md
PowerShell local AI	Install-Ollama.ps1, Pull-OllamaModels.ps1, Test-LocalAIEndpoint.ps1	PowerShell/local_ai/README.md
Perl developer utilities	jsonl_validate.pl, todo_report.pl, env_audit.pl, replace_in_files.pl, log_summary.pl, repo_lint.pl	Perl/README.md

---

Quick Start

Clone the repo

git clone https://github.com/mihaibc/automate_everything.git
cd automate_everything

Install development tools

python3 -m pip install -e ".[dev]"

Install optional AI dependencies

python3 -m pip install -e ".[ai]"

Count tokens before sending text to an LLM

echo "Your text here" | python Python/ai_utils/token_counter.py --model gpt-4

Set up a local LLM with Ollama

bash Bash/llm_setup/install_ollama.sh
bash Bash/llm_setup/pull_models.sh
bash Bash/llm_setup/run_inference.sh --prompt "Explain Kubernetes in one sentence"

Check local AI hardware signals

bash Bash/local_ai/check_gpu.sh

Organize files by extension safely

python Python/file_management/move_files.py \
  --source ~/Downloads \
  --destination ~/Documents/PDFs \
  --extensions pdf \
  --dry-run

Run batch LLM inference from a JSONL file

python Python/ai_utils/batch_inference.py \
  --input examples/prompts/prompts.jsonl \
  --output results.jsonl

---

Local AI Recipes

Inspect local acceleration options

bash Bash/local_ai/check_gpu.sh

Build llama.cpp

bash Bash/local_ai/install_llama_cpp.sh --accelerator auto --dry-run

Remove --dry-run when the planned clone and build path looks right.

Create a Python local-AI environment

bash Bash/local_ai/setup_python_ai_env.sh --venv .venv-local-ai

List installed Ollama models

python Python/local_ai/ollama_model_report.py

Benchmark an Ollama model

python Python/local_ai/benchmark_ollama.py --model llama3 --runs 3

Generate an Ollama Modelfile

python Python/local_ai/generate_modelfile.py \
  --from llama3 \
  --system "You are concise and practical." \
  --parameter "temperature 0.2"

Windows Ollama smoke test

PowerShell/local_ai/Test-LocalAIEndpoint.ps1 -Model llama3

---

Quality Gates

Run the same checks locally that CI runs:

make check

Individual commands:

make lint
make test
make shellcheck
make py-compile
make bash-test
make perl-syntax
make perl-test
make powershell-test
make docs-check
make script-test

Install shellcheck separately if your platform does not include it. PowerShell tests skip locally when pwsh is unavailable, but CI runs them.

---

Repository Standards

New scripts should include:

• A clear CLI with --help.
• Safe defaults and --dry-run for destructive operations.
• Quoted shell variables and strict Bash mode where relevant.
• No secrets, tokens, or personal paths.
• README usage examples.
• Tests for Python behavior when practical.
• A test mapping entry in tests/script_test_map.json.

---

Roadmap

• More local model runners: llama.cpp server mode, LM Studio, vLLM, and Open WebUI helpers.
• Cross-platform smoke tests for PowerShell scripts.
• More file-management helpers with dry-run and rollback support.
• Script metadata index for discoverability.
• First Go and C++ scripts after a concrete use case is ready.

Have a script that fits? Open a feature request or submit a PR.

---

Contributing

Contributions are welcome: new scripts, bug fixes, tests, examples, and documentation improvements. Please read CONTRIBUTING.md, SECURITY.md, and the Code of Conduct before opening a PR.

---

License

Apache-2.0. See LICENSE for details.