autopreso 0.1.1 → 0.1.5

package/README.md

@@ -29,13 +29,11 @@ Stage a few seed elements, hit start, and present.
 ```sh
 $ npx autopreso              # boots the server, opens the browser
 autopreso listening at http://127.0.0.1:3210
-whiteboard agent: openai gpt-5.5
-settings file: /Users/you/.config/autopreso/settings.json
 
 # In the browser:
 # 1. Drop reference materials onto the staging canvas (title, agenda, etc).
-# 2. Pick your microphone, pick a transcription model and an agent model.
-# 3. Click "Start preso" and start talking.
+# 2. Pick your microphone, transcription model, agent model, and optional Agent instructions.
+# 3. Click "Start Preso" and start talking.
 ```
 
 ## Install
@@ -80,10 +78,10 @@ npm start
                                                   └────────────────┘
 ```
 
-- **Two modes** - "staging" lets you sketch seed content client-side; "live" hands the canvas over to the agent and starts streaming transcripts.
+- **Two modes** - "staging" lets you sketch seed content client-side; "live" hands the canvas over to the agent, biases OpenAI Realtime transcription toward staging text and labels, and starts streaming transcripts.
 - **Local server, local network only** - the Express + WebSocket server binds to 127.0.0.1; nothing is exposed beyond your machine.
-- **Persistent settings** - models, API keys, and STT engine choices live in `~/.config/autopreso/settings.json` and survive restarts.
-- **Warmup loop** - after you hit start the agent primes itself against your staging content so the first sentence you say doesn't get a cold model.
+- **Persistent settings** - models, API keys, STT engine choices, and Agent instructions live in `~/.config/autopreso/settings.json` and survive restarts.
+- **Warmup loop** - after you hit start the agent primes itself against your staging content and Agent instructions so the first sentence you say doesn't get a cold model.
 
 ## CLI Reference
 
@@ -102,6 +100,9 @@ npm start
 ## Configuration
 
 Settings persist at `~/.config/autopreso/settings.json` and are managed from the in-app status panel.
+Agent instructions are saved automatically from staging, can be up to 100,000 characters, and take effect on the next Start Preso.
+The live Session cost card estimates agent token costs and OpenAI Realtime audio costs for the current presentation, resetting on Start Preso or session reset.
+OpenAI prices use the built-in May 2026 rate table; local providers show `$0.0000`, Codex shows token volume because it routes through your subscription, and unknown models show `n/a`.
 
 ### Defaults on first run
 
@@ -119,22 +120,24 @@ Auto-detection precedence: **Codex CLI auth wins over `OLLAMA_MODEL` wins over `
 
 ### Environment variables
 
-These only seed `settings.json` on first run. Once the file exists, they're ignored - edit the file or use the in-app panel.
+Provider variables only seed `settings.json` on first run. Once the file exists, they're ignored - edit the file or use the in-app panel. Log path variables are read on each process start.
 
-| Variable         | Purpose                                               |
-| ---------------- | ----------------------------------------------------- |
-| `PORT`           | Port to listen on. Default: `3210`.                   |
-| `OPENAI_API_KEY` | Seeds the OpenAI key for both agent and Realtime STT. |
-| `OPENAI_MODEL`   | Seeds the OpenAI agent model.                         |
-| `CODEX_MODEL`    | Seeds the Codex model.                                |
-| `OLLAMA_MODEL`   | Seeds the Ollama model.                               |
+| Variable               | Purpose                                               |
+| ---------------------- | ----------------------------------------------------- |
+| `PORT`                 | Port to listen on. Default: `3210`.                   |
+| `OPENAI_API_KEY`       | Seeds the OpenAI key for both agent and Realtime STT. |
+| `OPENAI_MODEL`         | Seeds the OpenAI agent model.                         |
+| `CODEX_MODEL`          | Seeds the Codex model.                                |
+| `OLLAMA_MODEL`         | Seeds the Ollama model.                               |
+| `AUTOPRESO_CACHE_LOG`  | Cache usage log path. Default: `~/.config/autopreso/logs/cache.log`. |
+| `AUTOPRESO_DEBUG_LOG`  | Agent debug log path. Default: `~/.config/autopreso/logs/debug.log`. |
 
 Local Moonshine transcription ships as an optional native sidecar for `darwin-arm64` and `darwin-x64`. On other platforms, choose OpenAI Realtime in the STT panel.
 
 ## Credits
 
 - [Excalidraw](https://github.com/excalidraw/excalidraw) - the whiteboard canvas, scene model, and rendering.
-- [Moonshine](https://github.com/usefulsensors/moonshine) by Useful Sensors - the local speech-to-text model that makes the offline path possible.
+- [Moonshine](https://github.com/moonshine-ai/moonshine) the local speech-to-text model that makes the offline path possible.
 - [Vercel AI SDK](https://github.com/vercel/ai) - tool-calling agent loop and provider abstraction.
 
 ## Development
@@ -142,6 +145,7 @@ Local Moonshine transcription ships as an optional native sidecar for `darwin-ar
 ```sh
 npm install                       # install deps
 npm run dev                       # run the CLI from source
+npm run typecheck                 # tsc --noEmit
 npm test                          # node --test
 npm run build:moonshine-sidecars  # build the Python sidecar binaries
 ```

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "autopreso",
-  "version": "0.1.1",
+  "version": "0.1.5",
   "description": "Realtime speech to presentation. Let the whiteboard whiteboard itself.",
   "license": "MIT",
   "author": "Kun Chen <kun@kunchenguid.com>",
@@ -41,7 +41,8 @@
     "dev": "node ./src/cli.js",
     "prepare:release-packages": "node ./scripts/prepare-release-packages.js",
     "test": "node --test",
-    "start": "node ./src/cli.js"
+    "start": "node ./src/cli.js",
+    "typecheck": "tsc --noEmit"
   },
   "dependencies": {
     "@ai-sdk/openai": "^3.0.63",
@@ -55,5 +56,10 @@
     "@autopreso/moonshine-darwin-arm64": "0.1.1",
     "@autopreso/moonshine-darwin-x64": "0.1.1"
   },
-  "devDependencies": {}
+  "devDependencies": {
+    "@types/express": "^5.0.6",
+    "@types/node": "^25.6.2",
+    "@types/ws": "^8.18.1",
+    "typescript": "^6.0.3"
+  }
 }