@mindstudio-ai/remy 0.1.343 → 0.1.345

package/README.md

@@ -360,10 +360,10 @@ All command responses include the `requestId` from the originating command.
 | `user_message` | `text`, `attachments?`, `queued?`, `hidden?` | Echo of a user message entering the turn. Queue-delivered messages (including ASAP items injected mid-turn) carry `queued: true` and their own original `requestId` (a merged turn emits one per absorbed message); idle sends echo with the turn's requestId and no `queued` flag. `hidden: true` marks internal entries (e.g. passive background results) that should not render. |
 | `tool_start` | `id`, `name`, `input`, `partial?`, `parentToolId?` | Tool execution started. `partial: true` means more `tool_start` events will follow for this id (progressive input streaming). |
 | `tool_input_delta` | `id`, `name`, `result`, `parentToolId?` | Progressive tool content (streaming tools only) |
-| `tool_done` | `id`, `name`, `result`, `isError`, `parentToolId?` | Tool execution completed. For a background tool this carries the synchronous ack; for a synthesized user/gate compaction block it arrives late, when the compaction finishes, carrying the summary (or error). |
+| `tool_done` | `id`, `name`, `result`, `isError`, `parentToolId?`, `recording?` | Tool execution completed. For a background tool this carries the synchronous ack; for a synthesized user/gate compaction block it arrives late, when the compaction finishes, carrying the summary (or error). `recording` is a browser-test replay chunk, lifted off the result string so the history cap cannot lose it. |
 | `tool_background_complete` | `id`, `name`, `result`, `parentToolId?` | A background tool's detached work finished; `result` belongs in the block's `backgroundResult`. Emitted for all `backgroundNotify` classes. |
 | `status` | `message` | Contextual status label (e.g., "Writing files...") |
-| `error` | `error` | Error message (may precede `completed`) |
+| `error` | `error`, `code?`, `badModelId?` | Error message (may precede `completed`). `code` is the server's structured classifier when it sent one; `badModelId` accompanies `code: 'invalid_model_override'` and names the rejected model. |
 | `history` | `messages` | Response to `get_history` |
 | `session_cleared` | | Response to `clear` |
 | `models_changed` | `models?`, `modelSurfaces`, `allowedModelsByType` | Response to `changeModels` |
@@ -435,9 +435,16 @@ npm run typecheck     # Type check only
 
 ## Config
 
-Remy reads credentials from `~/.mindstudio-local-tunnel/config.json`, using the active environment's `apiKey` and `apiBaseUrl`.
-
-Resolution order for API key:
+Resolution order for the API key:
 1. `--api-key` flag
 2. `MINDSTUDIO_API_KEY` environment variable
-3. `~/.mindstudio-local-tunnel/config.json` (active environment)
+
+And for the base URL: `--base-url`, then `MINDSTUDIO_BASE_URL`, then
+`https://api.mindstudio.ai`. `MINDSTUDIO_APP_ID` is env-only and optional — when unset the platform
+attributes cost to a shared per-org service-account app.
+
+Inside a dev box, the C&C server sets both variables when it spawns remy; there is nothing to
+configure. Running remy by hand means exporting them yourself.
+
+Remy used to fall back to `~/.mindstudio-local-tunnel/config.json`, written by the dev tunnel's
+login flow when the tunnel was a standalone CLI. Nothing writes that file any more.