@wes-dev/pr 0.0.1 → 0.0.3

package/CHANGELOG.md

@@ -7,6 +7,39 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.0.3] - 2026-07-13
+
+### Added
+
+- URL-first pull request loading with suffix-tolerant GitHub URL parsing and deep links to the selected logical group and file.
+- Editable and deletable inline draft comments stored with the in-memory review session.
+- Pull request description and review submission dialogs, including GitHub-flavored Markdown rendering.
+
+### Changed
+
+- Rebuilt the review workspace as a compact Windows 95-inspired interface using `css95`, with the package scale set to `1`.
+- Consolidated navigation into one collapsible, horizontally resizable sidebar with independently scrollable and vertically resizable group and file sections.
+- Moved grouping into the centered empty-state call to action and review submission into a focused dialog so the diff remains the primary workspace.
+- Split the web UI into focused header, sidebar, file-list, diff-workspace, dialog, routing, and annotation modules.
+
+### Fixed
+
+- Preserve the file selected by a deep review URL while loading its pull request session.
+- Clear selected diff lines after saving or cancelling an inline comment.
+- Prevent logical groups from overlapping or becoming mangled when the sidebar is constrained.
+- Correct compact header rendering in Safari and improve selected-file contrast in the file tree.
+
+## [0.0.2] - 2026-07-10
+
+### Added
+
+- Phase-specific LLM reasoning effort configuration for manifest planning, group agents, and the reducer, with JSON, environment, and UI controls.
+- Server-side LLM failure diagnostics and visible grouping fallback warnings.
+
+### Changed
+
+- Keep complete pull request diffs and chunk contents in server memory; review-session responses now contain metadata only, while the UI loads the selected file diff on demand.
+
 ## [0.0.1] - 2026-07-09
 
 ### Added

package/README.md

@@ -18,7 +18,7 @@ Pull request URLs work too:
 npx @wes-dev/pr https://github.com/owner/repo/pull/123
 ```
 
-Or open the local workspace and enter a PR number in the UI:
+Or open the local workspace and paste the pull request URL into the UI:
 
 ```sh
 npx @wes-dev/pr
@@ -26,7 +26,7 @@ npx @wes-dev/pr
 
 The CLI binds to `127.0.0.1` by default, tries port `7777` and increments until it finds a free port, prints the URL, and never opens a browser. `--host` and `--port` are available when needed.
 
-The package uses the locally installed [`gh`](https://cli.github.com/) CLI. Authenticate it first with `gh auth login`. The UI has separate Repository and PR number inputs; pasting a GitHub pull request URL into either one fills both fields. A blank repository falls back to the current working directory, and bare `npx @wes-dev/pr 123` remains available for that local-repository case. The tool deliberately does not list or search a large organization’s PRs.
+The package uses the locally installed [`gh`](https://cli.github.com/) CLI. Authenticate it first with `gh auth login`. The UI has a single URL input and accepts copied PR URLs with suffixes such as `/files`, `/changes`, or `/commits`, as well as query strings and fragments. CLI shorthand such as `owner/repo#123` and bare `npx @wes-dev/pr 123` remains available. The tool deliberately does not list or search a large organization’s PRs.
 
 ## Review flow
 
@@ -64,13 +64,20 @@ Configuration is read from `~/.config/wes-dev/pr/config.json`; environment varia
     "debug": false,
     "maxParallelAgents": 1,
     "maxRefinementRounds": 1,
-    "maxAgentIterations": 8
+    "maxAgentIterations": 8,
+    "reasoning": {
+      "manifest": "medium",
+      "groupAgent": "high",
+      "reducer": "medium"
+    }
   }
 }
 ```
 
 `omlx` is treated as an explicit OpenAI-compatible endpoint. For Ollama, use `preset: "ollama-openai"` and set `model`; its local OpenAI-compatible endpoint defaults to `http://127.0.0.1:11434/v1`.
 
+Set a reasoning field only when the configured runtime accepts the OpenAI Chat Completions `reasoning_effort` parameter. Omit a field to use the model default. Environment equivalents are `PR_LLM_REASONING_MANIFEST`, `PR_LLM_REASONING_GROUP_AGENT`, and `PR_LLM_REASONING_REDUCER`. UI changes apply to the current server process and are used by later grouping runs.
+
 ## Development and publishing
 
 ```sh