@blade-ai/orca 0.4.26-win32-x64 → 0.4.27-darwin-arm64

package/README.md

@@ -61,6 +61,12 @@ orca --resume [SESSION_ID]                # resume a saved conversation
 orca --fork SESSION_ID                    # fork a saved conversation
 ```
 
+Development builds also provide opt-in [shared ACP sessions](docs/acp-daemon.md)
+on Unix (`orca daemon`, `orca attach`, and `orca acp-bridge`),
+[file-defined subagents](docs/subagents.md), and
+[persistent terminal output pages](docs/architecture/adr/0006-unified-exec-terminal-service.md).
+These additions are available in v0.4.27 and later.
+
 On Windows PowerShell, set the key with `$env:DEEPSEEK_API_KEY = "sk-..."`;
 the `orca` commands are the same.
 
@@ -76,9 +82,10 @@ Orca prints the exact `orca --resume <SESSION_ID>` command for the session.
 Use `/plan` for read-only planning, `/goal` for a persistent objective, and
 `/tasks` for the unified workspace containing subagents, background commands,
 monitors, and workflow children. `/agents` is an alias for `/tasks`, while
-`/workflows` keeps the workflow-specific run tree. Use `/trust` to manage the
-current folder's sandbox permissions. Automatic project memory is enabled for
-recorded sessions by default; use `/remember` for explicit user or project
+`/workflows` keeps the workflow-specific run tree. Use `/trust` to control
+whether Orca loads project-provided configuration and instructions; it does not
+enable or bypass OS sandbox enforcement. Automatic project memory is enabled
+for recorded sessions by default; use `/remember` for explicit user or project
 facts. See [Memory](docs/memory.md) for capture, recall, storage, privacy, and
 deletion.
 
@@ -130,14 +137,17 @@ then defaults. Run `orca --help` or `orca exec --help` for the full command
 surface. User configuration lives at `~/.orca/config.toml`; trusted projects
 can also provide `.orca/config.toml`, `AGENTS.md`, rules, skills, and workflows.
 
-DeepSeek V4 thinking is enabled explicitly. Set `reasoning_effort` to `low`,
+DeepSeek thinking is enabled explicitly. Set `reasoning_effort` to `low`,
 `high`, or `max` (the default) in `config.toml`, or use
-`ORCA_REASONING_EFFORT`. `deepseek-v4-flash`, `deepseek-v4-flash-vision-exp`,
-and `deepseek-v4-pro` use a 1M-token context window and allow up to 384K output
-tokens. JPEG, PNG, GIF, and WebP inputs are accepted from ACP clients and the
-TUI with every model selection. The vision model consumes images directly;
-`auto`, Pro, and Flash use a task-aware vision analysis before continuing with
-the selected coding model. In the TUI, use `Ctrl+V` to attach the current
+`ORCA_REASONING_EFFORT`. `deepseek-flash` (DeepSeek-V4.1-Flash) and
+`deepseek-v4-pro` use a 1M-token context window and allow up to 384K output
+tokens. The retired `deepseek-v4-flash` and
+`deepseek-v4-flash-vision-exp` names remain accepted and normalize to
+`deepseek-flash`, including Flash pricing. JPEG, PNG, GIF, and WebP inputs are
+accepted from ACP clients and the TUI with every model selection. Explicit
+Flash selections consume images directly; `auto` and Pro use Flash for
+task-aware visual analysis before continuing with the selected coding model.
+In the TUI, use `Ctrl+V` to attach the current
 clipboard image (`Alt+V` is also available on Windows), drag or paste image
 paths and `file://` URLs, or select an image through `@file`. Each attachment
 appears as an atomic `[Image #N]` item that can be deleted, cleared, queued,

package/package.json

@@ -14,12 +14,12 @@
     "node": ">=16"
   },
   "name": "@blade-ai/orca",
-  "version": "0.4.26-win32-x64",
-  "description": "Native Orca binary for win32/x64.",
+  "version": "0.4.27-darwin-arm64",
+  "description": "Native Orca binary for darwin/arm64.",
   "os": [
-    "win32"
+    "darwin"
   ],
   "cpu": [
-    "x64"
+    "arm64"
   ]
 }