@counterpoint-studio/audio-file-mcp-app 1.0.1 → 1.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -5,12 +5,13 @@ for playing and inspecting local audio files in an MCP host.
5
5
 
6
6
  ![Audio File MCP App running in Claude Desktop](https://raw.githubusercontent.com/counterpoint-studio/audio-file-mcp-app/main/docs/screenshot.png)
7
7
 
8
- Renders an in-conversation UI with playback, metadata, loudness, and a
9
- spectrogram.
8
+ Renders an in-conversation UI with playback, metadata, loudness, a
9
+ spectrogram, and an optional annotation-lane timeline.
10
10
 
11
- File metadata, loudness statistics, the current playhead
12
- position, and any selected region are also exposed back to the model, so
13
- follow-up tasks can refer to what the user is actually hearing and looking at.
11
+ File metadata, loudness statistics, the current playhead position, any
12
+ selected region, and the annotation lanes active at the playhead are also
13
+ exposed back to the model, so follow-up tasks can refer to what the user is
14
+ actually hearing and looking at.
14
15
 
15
16
  ## Features
16
17
 
@@ -28,6 +29,13 @@ follow-up tasks can refer to what the user is actually hearing and looking at.
28
29
  - **Looping region selection.** Drag on the timeline to mark a region;
29
30
  playback loops over it, and the region's start/end are passed back to
30
31
  the model alongside the file's loudness and playhead state.
32
+ - **Annotation-lane timeline.** Supply timeline annotations and the widget
33
+ draws a stack of thin labelled lanes between the waveform and spectrogram,
34
+ aligned to the audio's own timeline. Each lane carries time spans, an
35
+ optional colour, and an optional envelope that fades span opacity along the
36
+ lane. Hovering a span reveals its label, and the lanes active at the
37
+ playhead — plus those starting, ending, or active within a selected region
38
+ — are reported back to the model.
31
39
 
32
40
  ## Install
33
41
 
@@ -54,6 +62,15 @@ Or add the server by hand to `claude_desktop_config.json`:
54
62
  }
55
63
  ```
56
64
 
65
+ ### Codex Desktop
66
+
67
+ 1. Go to Settings -> MCP Servers
68
+ 2. Add Server
69
+ 3. Name: `audiofile-mcp-app`
70
+ 4. Command to launch: `npx`
71
+ 5. Arguments `-y` and `@counterpoint-studio/audio-file-mcp-app`
72
+
73
+
57
74
  ### VS Code (Copilot, Agent mode)
58
75
 
59
76
  [![Install in VS Code](https://img.shields.io/badge/Install%20in-VS%20Code-007ACC?logo=visualstudiocode&logoColor=white)](https://insiders.vscode.dev/redirect/mcp/install?name=audio-file&config=%7B%22type%22%3A%22stdio%22%2C%22command%22%3A%22npx%22%2C%22args%22%3A%5B%22-y%22%2C%22%40counterpoint-studio%2Faudio-file-mcp-app%22%5D%7D)
@@ -100,18 +117,47 @@ example:
100
117
  The host calls the `display_audio_file` tool, which renders the in-app UI
101
118
  with waveform, spectrogram, loudness metrics, and playback transport.
102
119
 
120
+ ### Annotation lanes
121
+
122
+ `display_audio_file` accepts an optional `annotations` object describing lanes
123
+ to draw on the timeline (or an `annotationsPath` pointing at a JSON file with
124
+ the same `{ "lanes": [...] }` shape — handy for large payloads):
125
+
126
+ ```json
127
+ {
128
+ "annotations": {
129
+ "lanes": [
130
+ {
131
+ "label": "Delay tails",
132
+ "color": "#e6007e",
133
+ "spans": [{ "start": 8, "end": 24 }],
134
+ "envelope": [
135
+ { "time": 8, "value": 0 },
136
+ { "time": 24, "value": 1 }
137
+ ]
138
+ }
139
+ ]
140
+ }
141
+ }
142
+ ```
143
+
144
+ Times are in seconds on the audio's own timeline. `label`, `color`, and
145
+ `envelope` are optional; a lane with an envelope fades its span opacity along
146
+ the envelope curve, and uncoloured lanes use a light-accent fill. Overlapping
147
+ spans in a lane are truncated at the next span's start so rows never overlap.
148
+ The model can author annotations to point out sections, mark events, or
149
+ visualise an analysis it just ran.
150
+
103
151
  ## Client compatibility
104
152
 
105
153
  Tested and known to work in:
106
154
 
107
- - **Claude Desktop** — Chat and Cowork
155
+ - **Claude Desktop** — Chat, Cowork, and Code
156
+ - **Codex Desktop**
108
157
  - **Visual Studio Code**
109
158
  - **Goose**
110
159
  - **MCP Inspector**
111
160
 
112
- The Codex app has been tested too, but
113
- [its MCP App support is currently broken](https://github.com/openai/codex/issues/21019).
114
-
115
161
  ## Development
116
162
 
117
163
  ```bash