Release v0.2.0

Supervision interop (Detection.from_supervision), LineRule min_frames debounce,
ZoneRule grace/anchor robustness, two hero examples (rtsp_to_webhook,
from_supervision) + latent_vec, distribution renamed to trio-retina, complete
[all] extra, and an honesty pass on the forecast / latent-channel claims.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

Author: raullenchai

.github/ISSUE_TEMPLATE/bug_report.md

@@ -28,7 +28,7 @@ What actually happened (include the full traceback if there is one).
 
 ## Environment
 
-- Trio Retina version: <!-- e.g. 0.1.0 -->
+- Trio Retina version: <!-- e.g. 0.2.0 -->
 - Python version: <!-- e.g. 3.11 -->
 - OS: <!-- e.g. macOS 14, Ubuntu 22.04 -->
 - Optional extras installed: <!-- e.g. [yolo], [video], none -->

CHANGELOG.md

@@ -6,6 +6,8 @@ to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ## [Unreleased]
 
+## [0.2.0] — 2026-06-17
+
 ### Added
 
 - Example `examples/latent_vec.py`: populate the dual-state latent `vec` channel by
@@ -121,7 +123,8 @@ First public open-source release.
   (`JsonlSink`, `WebhookSink`).
 - `event_f1` / `match_events`: a generic metric to compare two event streams.
 
-[Unreleased]: https://github.com/machinefi/trio-retina/compare/v0.1.0...HEAD
+[Unreleased]: https://github.com/machinefi/trio-retina/compare/v0.2.0...HEAD
+[0.2.0]: https://github.com/machinefi/trio-retina/releases/tag/v0.2.0
 [0.1.0]: https://github.com/machinefi/trio-retina/releases/tag/v0.1.0
 [0.0.4]: https://github.com/machinefi/trio-retina
 [0.0.3]: https://github.com/machinefi/trio-retina

CITATION.cff

@@ -11,4 +11,4 @@ abstract: >-
   world models.
 repository-code: https://github.com/machinefi/trio-retina
 license: Apache-2.0
-version: 0.1.0
+version: 0.2.0

README.md

@@ -209,7 +209,7 @@ Full rationale, references, and the world-model stack: [`DESIGN.md`](DESIGN.md).
 
 ## 🗺️ roadmap
 
-Early but real (`v0.1.0`). Stable: the event layer + JSON Schema/validator, the composable pipeline (`|` / list / JSON), YOLO + open-vocab + VLM detectors (plus `from_supervision` interop), IoU + Norfair trackers, and jitter-robust rules (`exit_grace_s` · `anchor` · `min_frames`).
+Early but real (`v0.2.0`). Stable: the event layer + JSON Schema/validator, the composable pipeline (`|` / list / JSON), YOLO + open-vocab + VLM detectors (plus `from_supervision` interop), IoU + Norfair trackers, and jitter-robust rules (`exit_grace_s` · `anchor` · `min_frames`).
 
 Next: ByteTrack / OC-SORT · `proximity` / `anomaly` events · VLM-as-event-source · Kafka / MQTT sinks · the **latent channel** (surface V-JEPA scene + per-object embeddings). See [`CHANGELOG.md`](CHANGELOG.md).
 

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "trio-retina"
-version = "0.1.0"
+version = "0.2.0"
 description = "Model-agnostic computer-vision pipeline: turn any detector (YOLO, VLM, Grounding DINO) into structured tracking events — zone intrusion, line-crossing, dwell, counting — over video or RTSP. The state layer for world models."
 readme = "README.md"
 requires-python = ">=3.10"

retina/__init__.py

@@ -55,7 +55,7 @@
 from .worldstate import Entity, Relation, Vec, WorldState
 from .zones import Line, Zone
 
-__version__ = "0.1.0"
+__version__ = "0.2.0"
 
 __all__ = [
     # runner