determystic 0.1.0__tar.gz

determystic-0.1.0/.github/workflows/ci.yml

@@ -0,0 +1,46 @@
+name: CI
+
+on:
+  push:
+    branches: [ main ]
+  pull_request:
+    branches: [ main ]
+
+jobs:
+  lint:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+      
+      - name: Install uv
+        uses: astral-sh/setup-uv@v3
+      
+      - name: Install dependencies
+        run: uv sync
+      
+      - name: Run lint validation
+        run: make lint-validate
+
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.12'
+      
+      - name: Install uv
+        uses: astral-sh/setup-uv@v3
+      
+      - name: Install dependencies
+        run: uv sync
+      
+      - name: Run tests
+        run: make test

determystic-0.1.0/.gitignore

@@ -0,0 +1,210 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+#poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+#pdm.lock
+#pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+#pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.envrc
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# Abstra
+# Abstra is an AI-powered process automation framework.
+# Ignore directories containing user credentials, local state, and settings.
+# Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore 
+#  that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#  and can be added to the global gitignore or merged into this file. However, if you prefer, 
+#  you could uncomment the following to ignore the entire vscode folder
+# .vscode/
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Cursor
+#  Cursor is an AI-powered code editor. `.cursorignore` specifies files/directories to
+#  exclude from AI features like autocomplete and code analysis. Recommended for sensitive data
+#  refer to https://docs.cursor.com/context/ignore-files
+.cursorignore
+.cursorindexingignore
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/
+
+example_project
+.determystic

determystic-0.1.0/.python-version

@@ -0,0 +1 @@
+3.12

determystic-0.1.0/Makefile

@@ -0,0 +1,12 @@
+.PHONY: lint
+
+lint:
+	uv run ruff check --fix .
+	uv run ty check .
+
+lint-validate:
+	uv run ruff check .
+	uv run ty check .
+
+test:
+	uv run pytest -vvv

determystic-0.1.0/PKG-INFO

@@ -0,0 +1,114 @@
+Metadata-Version: 2.4
+Name: determystic
+Version: 0.1.0
+Summary: Add your description here
+Requires-Python: >=3.12
+Requires-Dist: anthropic>=0.39.0
+Requires-Dist: click>=8.1.0
+Requires-Dist: pydantic-ai>=0.0.14
+Requires-Dist: pydantic-settings>=2.0.0
+Requires-Dist: pytest-asyncio>=0.24.0
+Requires-Dist: pytest>=8.0.0
+Requires-Dist: pyyaml>=6.0.0
+Requires-Dist: rich>=13.0.0
+Requires-Dist: ruff>=0.7.0
+Requires-Dist: tomli-w>=1.0.0
+Requires-Dist: ty
+Description-Content-Type: text/markdown
+
+# determystic
+
+Determystic is a library that forces your agents to give you a coding style that you're happy with - _deterministically_ every time.
+
+It works by creating validators for your coding conventions, using the AST of your programming language. If you see a bad piece of code and can describe why it's bad and why you never want to see code like that in the future, there's a good chance we can write a deterministic validator to make sure it never happens again.
+
+## Getting Started
+
+We'll look into adding a MCP server in the future. But for the time being just asking Claude Code / Cursor to run our validation script is usually good enough to get the job done.
+
+Append to your `.cursorrules` (Cursor) or `CLAUDE.md` (Claude Code).
+
+```
+Before yielding results, you should ALWAYS call `uvx determystic validate`.
+```
+
+Also setup your anthropic key for our global configuration. This will be accessible across projects so you should only have to do this once:
+
+```bash
+uvx determystic configure
+```
+
+When you have an issue, you can add a special validation case using:
+
+```bash
+uvx determystic new-validator
+```
+
+## Example
+
+Let's say your LLM generated some code that we don't like:
+
+```bash
+from typing import Optional
+from pydantic import BaseModel
+
+
+class MyModel(BaseModel):
+    name: Optional[str] = None
+    age: int
+
+
+def main():
+    model = MyModel(name="John", age=30)
+    print(model)
+
+if __name__ == "__main__":
+    main()
+```
+
+In this case, the fact that there's an Optional typehint instead of modern Python syntax (A | None = None). We can add this as a rule:
+
+```bash
+Code: name: Optional[str] = None
+Feedback: Don't use Optional - use A | None
+```
+
+This will add a new .deterministic hidden folder in your current project that you can introspect. But usually the logic looks pretty good zero-shot (we add internal tests to try to ensure that reasonableness of the validator we just wrote) so we can then run the validation:
+
+```bash
+$ uvx determystic validate example_project
+
+Detailed Results:
+
+✗ Custom Validator
+  main.py:6: Use 'T | None' instead of 'Optional[T]' for type hints
+        4 | 
+        5 | class MyModel(BaseModel):
+  >>>   6 |     name: Optional = None
+        7 |     age: int
+        8 |
+```
+
+## Background
+
+Programming agents are getting _really good_. You're hard pressed to find a professional engineer these days that doesn't use Cursor or Claude Code for at least some part of their workflow.
+
+My main annoyance in using these systems is when they output code that mostly works but is really messy, or against my own coding conventions. Typehinting in Python is especially egregious here. No matter how much I try to coerce my AGENT.md files, all of the SOTA models have a very strong preference to use List[] and Optional[]. I want to use the modern `list[]` and `A | None`.
+
+It's a small thing but it's representative of a larger problem. The main control we have today over these systems today is in their system prompts: specifying a AGENT.md or .cursorrules file to try to guide their behavior over text alone. This certainly works for higher level instructions like describing a feature scope. But we lose precision over what we're looking for by having to describe programming goals and constructs in natural language instead of code. Adding in AST validation changes that - and it turns out that LLMs are actually very good at writing AST validators even though they're pretty annoying for people
+
+## How it works
+
+When you see your LLMs outputting something that you know you never want in practice, you'll want to add a deterministic "validator". Copy some portion of your code file that exhibits the problem and run:
+
+```bash
+uv run ...
+```
+
+## Random notes
+
+- Targeting just Python for now. Other languages can follow the same convention pretty closely, but we need to support AST validating for their syntax & test whether LLMs will output better AST validators when written in the same language or if we can use Python as a bridge for control logic
+- Using Anthropic's Claude to do the authoring of the AST validators and the testing files (although in theory it would be very easy to swap this out for any other coding model)
+- We use .deterministic file extensions for our validation and validation test files. These are just python files but we prefer a different extension so they're not inadvertantly picked up by static analysis tools that just sniff for any .py extension. We might reconsider this in the future.
+- Since determystic files are on disk, they should be portable across projects and usable by CI validation across a team
+- Right now we don't support the editing case for existing validators - but this seems like an obvious extension in the future to try and make these more flexible given additional code that either incorrectly validates or does not validate

determystic-0.1.0/README.md

@@ -0,0 +1,96 @@
+# determystic
+
+Determystic is a library that forces your agents to give you a coding style that you're happy with - _deterministically_ every time.
+
+It works by creating validators for your coding conventions, using the AST of your programming language. If you see a bad piece of code and can describe why it's bad and why you never want to see code like that in the future, there's a good chance we can write a deterministic validator to make sure it never happens again.
+
+## Getting Started
+
+We'll look into adding a MCP server in the future. But for the time being just asking Claude Code / Cursor to run our validation script is usually good enough to get the job done.
+
+Append to your `.cursorrules` (Cursor) or `CLAUDE.md` (Claude Code).
+
+```
+Before yielding results, you should ALWAYS call `uvx determystic validate`.
+```
+
+Also setup your anthropic key for our global configuration. This will be accessible across projects so you should only have to do this once:
+
+```bash
+uvx determystic configure
+```
+
+When you have an issue, you can add a special validation case using:
+
+```bash
+uvx determystic new-validator
+```
+
+## Example
+
+Let's say your LLM generated some code that we don't like:
+
+```bash
+from typing import Optional
+from pydantic import BaseModel
+
+
+class MyModel(BaseModel):
+    name: Optional[str] = None
+    age: int
+
+
+def main():
+    model = MyModel(name="John", age=30)
+    print(model)
+
+if __name__ == "__main__":
+    main()
+```
+
+In this case, the fact that there's an Optional typehint instead of modern Python syntax (A | None = None). We can add this as a rule:
+
+```bash
+Code: name: Optional[str] = None
+Feedback: Don't use Optional - use A | None
+```
+
+This will add a new .deterministic hidden folder in your current project that you can introspect. But usually the logic looks pretty good zero-shot (we add internal tests to try to ensure that reasonableness of the validator we just wrote) so we can then run the validation:
+
+```bash
+$ uvx determystic validate example_project
+
+Detailed Results:
+
+✗ Custom Validator
+  main.py:6: Use 'T | None' instead of 'Optional[T]' for type hints
+        4 | 
+        5 | class MyModel(BaseModel):
+  >>>   6 |     name: Optional = None
+        7 |     age: int
+        8 |
+```
+
+## Background
+
+Programming agents are getting _really good_. You're hard pressed to find a professional engineer these days that doesn't use Cursor or Claude Code for at least some part of their workflow.
+
+My main annoyance in using these systems is when they output code that mostly works but is really messy, or against my own coding conventions. Typehinting in Python is especially egregious here. No matter how much I try to coerce my AGENT.md files, all of the SOTA models have a very strong preference to use List[] and Optional[]. I want to use the modern `list[]` and `A | None`.
+
+It's a small thing but it's representative of a larger problem. The main control we have today over these systems today is in their system prompts: specifying a AGENT.md or .cursorrules file to try to guide their behavior over text alone. This certainly works for higher level instructions like describing a feature scope. But we lose precision over what we're looking for by having to describe programming goals and constructs in natural language instead of code. Adding in AST validation changes that - and it turns out that LLMs are actually very good at writing AST validators even though they're pretty annoying for people
+
+## How it works
+
+When you see your LLMs outputting something that you know you never want in practice, you'll want to add a deterministic "validator". Copy some portion of your code file that exhibits the problem and run:
+
+```bash
+uv run ...
+```
+
+## Random notes
+
+- Targeting just Python for now. Other languages can follow the same convention pretty closely, but we need to support AST validating for their syntax & test whether LLMs will output better AST validators when written in the same language or if we can use Python as a bridge for control logic
+- Using Anthropic's Claude to do the authoring of the AST validators and the testing files (although in theory it would be very easy to swap this out for any other coding model)
+- We use .deterministic file extensions for our validation and validation test files. These are just python files but we prefer a different extension so they're not inadvertantly picked up by static analysis tools that just sniff for any .py extension. We might reconsider this in the future.
+- Since determystic files are on disk, they should be portable across projects and usable by CI validation across a team
+- Right now we don't support the editing case for existing validators - but this seems like an obvious extension in the future to try and make these more flexible given additional code that either incorrectly validates or does not validate

determystic-0.1.0/determystic/__tests__/agents/__init__.py

@@ -0,0 +1 @@
+# Test package