Initial commit

.aider.chat.history.md

@@ -0,0 +1,5 @@
+
+# aider chat started at 2026-06-16 22:37:03
+
+> No git repo found, create one to track aider's changes (recommended)? (Y)es/(N)o [Yes]: n  
+> /home/sirian/.local/bin/aider  

.aider.conf.yml

@@ -0,0 +1,15 @@
+model: openrouter/deepseek/deepseek-v4
+weak-model: openrouter/qwen/qwen3-coder
+editor-model: openrouter/qwen/qwen3-coder
+edit-format: diff
+auto-commits: true
+dirty-commits: false
+auto-test: true
+test-cmd: "pytest tests/ -x"
+read:
+  - SPEC.md
+  - ARCH.md
+  - CONVENTIONS.md
+  - TESTS.md
+gitignore: true
+show-model-warnings: true

.aider.input.history

@@ -0,0 +1,3 @@
+
+# 2026-06-16 22:37:09.639627
++n

.gitignore

@@ -0,0 +1,31 @@
+.env
+.env.local
+*.key
+*.pem
+
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+venv/
+.venv/
+env/
+ENV/
+.pytest_cache/
+.mypy_cache/
+.ruff_cache/
+.coverage
+htmlcov/
+dist/
+build/
+*.egg-info/
+
+# IDE
+.vscode/
+
+*.log
+logs/
+
+build/

ARCH.md

@@ -0,0 +1,34 @@
+# ARCH - [project name]
+
+Architectural decisions and their reasoning. What components are used, how will they communicate, what libraries are to be used...
+
+## Overview
+
+Short description of structure. Diagrams and pictures are helpful.
+
+## Tech stack
+
+Brief description of technologies to be used.
+
+For example:
+- Pyton, 3.11+, Async support required
+- Pytest, 8.x, Testing
+
+## Most important architectural decisions.
+
+Examples:
+
+### Polling vs events
+- Decision: Polling every 10s
+- Options: Using interrupts
+- Why?: More simple to implement.
+
+## File structure
+
+src/
+  main.py
+  config.py
+
+tests/
+  tests.py
+

CONVENTIONS.md

@@ -0,0 +1,23 @@
+# CONVENTIONS
+
+Rules ai tools must follow
+
+## General
+
+- Write code you understand yourself. If you cannot explain a line it doesn't belong here
+- No "clever" one-liners if a longer form is clearer
+- Function name says what it does, docstring why
+- No comments that restate the code. Comments say "why" not "what"
+
+## Git
+
+- One logical change per commit
+- Commit message in active voice: "add feature x" NOT "added feature X"
+
+
+## What NOT to do
+
+- No new dependencies without discussion (propose to ARCH.md, don't add)
+- Don't mess with code style. Follow what is already there
+- Don't refactor adjacent code "while you're at it"
+- No assumed environment variables without decumenting in ARCH.md

JOURNAL.md

@@ -0,0 +1,7 @@
+# JOURNAL
+
+Write a couple of sentences after every session. NOT AI GENERATED. Forces you to articulate what was done and helps to restore context after returning to the project a week later.
+
+Format: date + what was done + why + what's next
+
+## YYYY-MM-DD

README.md

@@ -0,0 +1,13 @@
+# Project name
+
+A brief description of what this project does.
+
+## Setup
+
+How does one run this project
+
+## Project structure
+
+Descriptions of files and structures.
+
+## Workflow

SPEC.md

@@ -0,0 +1,35 @@
+# SPEC - [project name]
+
+This file defines WHAT the project does. 
+
+## Goal
+
+One or two sentences about what this project will solve and for who. 
+
+## Acceptance criteria
+
+Project is ready when following concrete criteria are met and tested:
+
+- [ ] [criteria 1]
+- [ ] [criteria 2]
+
+## Constraints
+
+Technical boundaries and limitations:
+
+- Constraint 1
+- Constraint 2
+
+## Out of scope
+
+What is not part of this project:
+
+- Out of scope 1
+- Out of scope 2
+
+## Invariants
+
+Things that must never be broken:
+
+- Invariant 1
+- Invariant 2

TESTS.md

@@ -0,0 +1,4 @@
+# TESTS
+
+Test cases written in my own words. AI tools should read and implement tests written here in natural language.
+