docs: professional README and OSS layout for v0.2.0

.gitignore

@@ -1,31 +1,36 @@
 # Python-generated files
 __pycache__/
 *.py[oc]
+*.pyc
 build/
 dist/
 wheels/
-*.egg-info
+*.egg-info/
 
 # Virtual environments
-.venv
-.venv-ship
+.venv/
+.venv-ship/
 .env
 
+# Secrets / sessions
+.current_session
+
+# Local runtime data
 qdrant_data/
 repos/
 graphify-out/
-.env
-# Generated repository intelligence
-.codeturtle/
-
-# Local DB
+tmp/
 data/
+outputs/
 
-# Sessions
-.current_session
+# Snapshots / artifacts
+artifacts/last_review_snapshot.json
+
+# Codeturtle generated intelligence
+.codeturtle/
 
-# Debug
+# Debug text dumps
 debug_*.txt
 
-# Eval snapshots from live reviews
-artifacts/
+# GitHub
+.github/copilot-instructions.md

CHANGELOG.md

@@ -0,0 +1,7 @@
+# Changelog
+
+## 0.2.0
+
+- Installable `codeturtle` console script
+- Smoke: confident-ai/deepeval#3288 --dry-run → REQUEST_CHANGES
+- Offline goldens: qw-538, qw-571, deepeval-3288

CONTRIBUTING.md

@@ -0,0 +1,29 @@
+# Contributing
+
+## Setup
+
+Python 3.11+.
+
+```bash
+uv sync
+pip install -e ".[ollama,graphify]"
+```
+
+`codeturtle --help` must work.
+
+## Checks before PR
+
+```bash
+codeturtle --help
+uv run python -m tests.evaluation.run_eval --offline
+```
+
+Both must pass. Do not commit `.env` or tokens.
+
+## Reviews
+
+Default is `--dry-run`. Do not `--comment` on repos you do not own.
+
+## Issues
+
+Paste `--dry-run -v` logs with tokens redacted.