diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
index dd410f0..5154e6b 100644
--- a/.github/workflows/tests.yml
+++ b/.github/workflows/tests.yml
@@ -22,7 +22,8 @@ jobs:
         run: ruff check .
 
   test:
-    needs: lint
+    # Runs independently of lint: a style regression must not suppress the test
+    # signal. Merge-gating for both jobs is enforced by branch protection.
     runs-on: ubuntu-latest
     strategy:
       matrix:
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 2872ce6..b215a03 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -5,11 +5,13 @@ All notable changes to spatialtissuepy are documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
-## [Unreleased]
+## [0.3.0] - 2026-07-23
 
 PhysiCell reader correctness fixes, reported against v0.2.0 by the
 llm-abm-consistency harness (OHSU ChangLab) and reproduced against both their
-PhysiCell 1.14.2 output and this repository's bundled example simulation.
+PhysiCell 1.14.2 output and this repository's bundled example simulation. This
+release also repairs the CI lint gate, which had never passed, and raises the
+Python floor to match what the dependencies actually require.
 
 ### Fixed
 
@@ -70,12 +72,39 @@ PhysiCell 1.14.2 output and this repository's bundled example simulation.
 - `PhysiCellTimeStep._load_cell_data()` and `to_trajectory_dataframe()` now
   resolve columns from each frame's own XML, so models with differing variable
   counts parse correctly in the same session.
+- **Raised the minimum Python version to 3.10** (from a nominal 3.8).
+  `requires-python` claimed `>=3.8`, but the `mcp` extra depends on `fastmcp`,
+  which requires 3.10+, so `pip install spatialtissuepy[all]` could never
+  succeed on 3.8 or 3.9. The two had drifted since the MCP server was added and
+  nothing surfaced it, because the CI lint gate failed before any install ran.
+  Python 3.8 reached end of life in October 2024 and 3.9 in October 2025, and
+  current releases of the core scientific stack already require 3.11+. The CI
+  matrix now runs 3.10-3.13.
+
+### Build
+
+- **The `ruff` lint job passes for the first time.** It reported ~6,100 errors
+  on every branch, and because the test job declares `needs: lint`, no pull
+  request had been able to run the test suite. The linter was also flagging
+  real defects behind the noise: `Union` used in annotations but never imported
+  across seven modules (latent under `from __future__ import annotations`, but
+  raised `NameError` under `typing.get_type_hints`), a matplotlib symbol
+  referenced in module-level annotations while imported only lazily, and three
+  broken forward references. The bulk of the remaining diff was trailing
+  whitespace, import ordering, and unused locals.
+- Moved the deprecated top-level `ruff` config under `[tool.ruff.lint]` and
+  aligned the `black`, `ruff`, and `mypy` target versions with the new floor.
+- **Decoupled the `lint` and `test` CI jobs.** The test job no longer declares
+  `needs: lint`, so the two run as independent siblings. Previously a lint
+  failure skipped the entire test matrix, meaning a style regression could
+  silently suppress all test signal — which is how the `fastmcp` install
+  incompatibility went undetected.
 
 ### Known gaps
 
 - `parse_microenvironment_mat` is implemented and exported but still not wired
   into `PhysiCellTimeStep` / `PhysiCellSimulation`; substrate fields remain
-  reachable only through the parser. Planned for v0.3.0.
+  reachable only through the parser. Planned for a future release.
 
 ## [0.2.0] - 2026-06-26
 
@@ -118,4 +147,5 @@ numerical-stability improvements across the analysis modules.
 - Spatial and statistics bug fixes and numerical-stability improvements.
 - Runtime bugs in the MCP synthetic and viz tools.
 
+[0.3.0]: https://github.com/emcramer/spatialtissuepy/compare/v0.2.0...v0.3.0
 [0.2.0]: https://github.com/emcramer/spatialtissuepy/releases/tag/v0.2.0
diff --git a/pyproject.toml b/pyproject.toml
index 8d316c3..b96b751 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "spatialtissuepy"
-version = "0.2.0"
+version = "0.3.0"
 description = "Spatial analysis tools for tissue biology and multiplexed imaging"
 readme = "README.md"
 license = {text = "MIT"}
diff --git a/tests/test_physicell_labels.py b/tests/test_physicell_labels.py
index a47a58a..9681ec1 100644
--- a/tests/test_physicell_labels.py
+++ b/tests/test_physicell_labels.py
@@ -1,5 +1,5 @@
 """
-Regression tests for PhysiCell reader defects fixed in v0.2.1.
+Regression tests for PhysiCell reader defects fixed in v0.3.0.
 
 These cover the label-driven column extraction and orientation handling
 reported by the llm-abm-consistency harness (OHSU ChangLab, 2026-07-21).