Spring 2026

Author: denisecase

.editorconfig

@@ -0,0 +1,119 @@
+# ============================================================
+# .editorconfig (Keep files consistent across editors and IDEs)
+# ============================================================
+# SOURCE: https://github.com/denisecase/templates
+
+# REQ.UNIVERSAL: All professional GitHub project repositories MUST include .editorconfig.
+# WHY: Establish a cross-editor baseline so diffs stay clean and formatting is consistent.
+# ALT: Repository may omit .editorconfig ONLY if formatting is enforced equivalently by CI and formatter tooling.
+# CUSTOM: Adjust indent_size defaults only if organizational standards change; keep stable across projects.
+# NOTE: Sections are ordered by editorial importance, not strict alphabetical order.
+# EditorConfig is documented at https://editorconfig.org
+
+root = true
+
+
+# === Global defaults (always apply) ===
+
+[*]
+# WHY: Normalize line endings and encoding across Windows, macOS, and Linux.
+end_of_line = lf
+charset = utf-8
+
+# WHY: Newline at EOF avoids noisy diffs and tool warnings.
+insert_final_newline = true
+
+# WHY: Remove accidental whitespace noise in diffs.
+trim_trailing_whitespace = true
+
+# WHY: Default to 2 spaces for configs and markup; language-specific overrides follow.
+indent_style = space
+indent_size = 2
+
+
+# === Build systems (special rules) ===
+
+[Makefile]
+# WHY: Makefiles require tabs.
+indent_style = tab
+
+[*.mk]
+# WHY: Makefile includes require tabs.
+indent_style = tab
+
+
+# === Markup and documentation ===
+
+[*.md]
+# WHY: Keep Markdown clean; use explicit <br> for hard line breaks.
+indent_size = 2
+trim_trailing_whitespace = true
+
+[*.{tex,cls,sty}]
+# WHY: LaTeX convention is 2 spaces.
+indent_size = 2
+indent_style = space
+
+[*.xml]
+# WHY: XML convention is 2 spaces.
+indent_size = 2
+indent_style = space
+
+[*.{yml,yaml}]
+# WHY: YAML convention is 2 spaces.
+indent_size = 2
+indent_style = space
+
+
+# === Data and configuration ===
+
+[*.{json,jsonc,jsonl,ndjson}]
+# WHY: JSON tooling typically expects 2 spaces.
+indent_size = 2
+indent_style = space
+
+[*.toml]
+# WHY: TOML often follows 4-space indentation in many projects.
+indent_size = 4
+indent_style = space
+
+
+# === Programming languages ===
+
+[*.{js,ts}]
+# WHY: JS/TS ecosystem commonly uses 2 spaces.
+indent_size = 2
+indent_style = space
+
+[*.{py,pyi}]
+# WHY: Python convention is 4 spaces.
+indent_size = 4
+indent_style = space
+
+[*.ps1]
+# WHY: PowerShell convention is 4 spaces.
+indent_size = 4
+indent_style = space
+
+[*.{c,cpp,h,java,cs,go,rs}]
+# WHY: Many C-family and systems languages commonly use 4 spaces.
+indent_size = 4
+indent_style = space
+
+[*.{sh,bash}]
+# WHY: Shell script convention is 2 spaces.
+indent_size = 2
+indent_style = space
+
+
+# === Proof assistants and formal languages ===
+
+[*.lean]
+# WHY: Lean 4 convention is 2 spaces; matches Mathlib and stdlib style.
+indent_size = 2
+indent_style = space
+
+[*.{v,vo}]
+# WHY: Coq convention is 2 spaces.
+indent_size = 2
+indent_style = space

.gitattributes

@@ -0,0 +1,131 @@
+# ============================================================
+# .gitattributes (Keep files consistent across operating systems)
+# ============================================================
+# SOURCE: https://github.com/denisecase/templates
+
+# REQ.UNIVERSAL: All professional GitHub project repositories MUST include .gitattributes.
+# WHY: Ensure consistent line endings, diff behavior, and file classification
+# across Windows, macOS, and Linux environments.
+# ALT: Repository may omit .gitattributes ONLY if equivalent normalization is
+# enforced reliably by tooling and CI (rare and fragile).
+# CUSTOM: Update file-type rules only when introducing new languages,
+# binary artifacts, or documentation formats.
+# NOTE: Rules are ordered by impact and generality, not alphabetically.
+# Git attributes are documented at https://git-scm.com/docs/gitattributes
+
+# === Core defaults (always apply) ===
+
+# WHY: Auto-detect text files and normalize line endings to avoid cross-platform drift.
+* text=auto
+
+
+# === Programming languages and scripts ===
+
+# WHY: Python and shell scripts must use LF for CI/CD, Linux environments, and containers.
+*.py    text eol=lf
+*.sh    text eol=lf
+
+# WHY: PowerShell convention on Windows uses CRLF.
+*.ps1   text eol=crlf
+
+
+# === Markup and documentation ===
+
+# WHY: Documentation and markup files use LF; standard for cross-platform tooling.
+*.md    text eol=lf
+*.tex   text eol=lf
+*.sty   text eol=lf
+*.cls   text eol=lf
+*.bib   text eol=lf
+
+
+# === Configuration and structured text ===
+
+# WHY: Configuration and structured text formats use LF for stable diffs.
+*.json     text eol=lf
+*.jsonc   text eol=lf
+*.jsonl   text eol=lf
+*.ndjson  text eol=lf
+*.toml    text eol=lf
+*.yaml    text eol=lf
+*.yml     text eol=lf
+
+
+# === Proof assistants and formal languages ===
+
+# WHY: Lean source files must use LF for cross-platform consistency and CI.
+*.lean   text eol=lf
+
+# WHY: Lean build artifacts are binary; prevent normalization and meaningless diffs.
+*.olean  binary
+*.ilean  binary
+*.trace  binary
+
+# WHY: Coq source uses LF; compiled objects are binary.
+*.v      text eol=lf
+*.vo     binary
+*.vok    binary
+*.vos    binary
+*.glob   binary
+
+# WHY: Lake build directory should be excluded, but if tracked, treat as binary.
+.lake/**  binary
+
+
+# === Notebooks ===
+
+# WHY: Jupyter notebooks require specialized diff and merge handling.
+*.ipynb   diff=jupyternotebook
+*.ipynb   merge=jupyternotebook
+
+
+# === Databases (binary) ===
+
+# WHY: Database files are binary; prevent text normalization and meaningless diffs.
+*.db       binary
+*.duckdb   binary
+*.sqlite   binary
+*.sqlite3  binary
+
+
+# === Columnar and analytical data (binary) ===
+
+# WHY: Columnar and analytical data formats are binary; diffs are not meaningful.
+*.arrow    binary
+*.avro     binary
+*.feather  binary
+*.orc      binary
+*.parquet  binary
+
+
+# === Office, BI, PDFs, and compressed artifacts (binary) ===
+
+# WHY: Office documents, BI files, PDFs, and compressed artifacts are binary.
+*.7z     binary
+*.bz2    binary
+*.docx   binary
+*.gz     binary
+*.pbix   binary
+*.pbit   binary
+*.pdf    binary
+*.pptx   binary
+*.rar    binary
+*.tar    binary
+*.tgz    binary
+*.xls    binary
+*.xlsm   binary
+*.xlsx   binary
+*.xz     binary
+*.zip    binary
+
+
+# === GitHub metadata and UI ===
+
+# WHY: Exclude documentation and tests from GitHub language statistics.
+docs/**    linguist-documentation
+tests/**   linguist-documentation
+
+# === LOG FILES ===
+# WHY: Log files are typically large and binary in nature; prevent text normalization.
+*.log binary
+project.log binary

.github/AGENTS.md

@@ -0,0 +1,68 @@
+# AI Instructions (AGENTS / CLAUDE / Copilot)
+
+## WHY
+
+- This repo uses a uniform, reproducible workflow based on **uv** and **pyproject.toml**.
+- These instructions exist to prevent tool drift (e.g., pip) and OS mismatch.
+
+## Requirements
+
+- Use **uv** for all environment, dependency, and run commands in this repo.
+- Do **not** recommend or use `pip install ...` as the primary workflow.
+- This repo targets **Python 3.14**, pinned via uv.
+- Commands and guidance must work on Windows, macOS, and Linux.
+- If shell-specific commands are unavoidable, provide both:
+  - PowerShell (Windows)
+  - bash/zsh (macOS/Linux)
+
+## Quickstart
+
+- Install **uv** using the official method for your OS.
+- Keep uv current.
+- Pin Python 3.14 for this project using uv.
+- Sync dependencies (dev + docs) and upgrade.
+
+```shell
+uv self update
+uv python pin 3.14
+uv sync --extra dev --extra docs --upgrade
+```
+
+## pre-commit
+
+- pre-commit runs only on tracked / staged files.
+- Developers should `git add -A` files before expecting hooks to run.
+- Developers may need to **repeat** the git add / git commit commands if files changed
+
+```
+uvx pre-commit install
+git add -A
+uvx pre-commit run --all-files
+```
+
+## Run Python as a Module (src/nlp/)
+
+```
+uv run python -m nlp.module_name
+```
+
+# Notebooks in notebooks/
+
+Select the kernel and Run All.
+
+## Common Tasks
+
+Run all commands via **uv**.
+
+Lint / format:
+
+```shell
+uv run ruff format .
+uv run ruff check .
+```
+
+Build documentation:
+
+```shell
+uv run zensical build
+```

.github/copilot-instructions.md

@@ -0,0 +1,68 @@
+# AI Instructions (AGENTS / CLAUDE / Copilot)
+
+## WHY
+
+- This repo uses a uniform, reproducible workflow based on **uv** and **pyproject.toml**.
+- These instructions exist to prevent tool drift (e.g., pip) and OS mismatch.
+
+## Requirements
+
+- Use **uv** for all environment, dependency, and run commands in this repo.
+- Do **not** recommend or use `pip install ...` as the primary workflow.
+- This repo targets **Python 3.14**, pinned via uv.
+- Commands and guidance must work on Windows, macOS, and Linux.
+- If shell-specific commands are unavoidable, provide both:
+  - PowerShell (Windows)
+  - bash/zsh (macOS/Linux)
+
+## Quickstart
+
+- Install **uv** using the official method for your OS.
+- Keep uv current.
+- Pin Python 3.14 for this project using uv.
+- Sync dependencies (dev + docs) and upgrade.
+
+```shell
+uv self update
+uv python pin 3.14
+uv sync --extra dev --extra docs --upgrade
+```
+
+## pre-commit
+
+- pre-commit runs only on tracked / staged files.
+- Developers should `git add -A` files before expecting hooks to run.
+- Developers may need to **repeat** the git add / git commit commands if files changed
+
+```
+uvx pre-commit install
+git add -A
+uvx pre-commit run --all-files
+```
+
+## Run Python as a Module (src/nlp/)
+
+```
+uv run python -m nlp.module_name
+```
+
+# Notebooks in notebooks/
+
+Select the kernel and Run All.
+
+## Common Tasks
+
+Run all commands via **uv**.
+
+Lint / format:
+
+```shell
+uv run ruff format .
+uv run ruff check .
+```
+
+Build documentation:
+
+```shell
+uv run zensical build
+```