@ghostspeak/plugin-gateway-ghost 0.1.0
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 +94 -0
- package/dist/index.js +11433 -0
- package/dist/index.js.map +116 -0
- package/dist/src/config.d.ts +56 -0
- package/dist/src/config.d.ts.map +1 -0
- package/dist/src/gateway.d.ts +34 -0
- package/dist/src/gateway.d.ts.map +1 -0
- package/dist/src/handlers/image-description.d.ts +15 -0
- package/dist/src/handlers/image-description.d.ts.map +1 -0
- package/dist/src/handlers/image-generation.d.ts +15 -0
- package/dist/src/handlers/image-generation.d.ts.map +1 -0
- package/dist/src/handlers/index.d.ts +12 -0
- package/dist/src/handlers/index.d.ts.map +1 -0
- package/dist/src/handlers/object-generation.d.ts +20 -0
- package/dist/src/handlers/object-generation.d.ts.map +1 -0
- package/dist/src/handlers/text-embedding.d.ts +12 -0
- package/dist/src/handlers/text-embedding.d.ts.map +1 -0
- package/dist/src/handlers/text-generation.d.ts +34 -0
- package/dist/src/handlers/text-generation.d.ts.map +1 -0
- package/dist/src/handlers/text-to-speech.d.ts +12 -0
- package/dist/src/handlers/text-to-speech.d.ts.map +1 -0
- package/dist/src/handlers/tokenizer.d.ts +16 -0
- package/dist/src/handlers/tokenizer.d.ts.map +1 -0
- package/dist/src/handlers/transcription.d.ts +12 -0
- package/dist/src/handlers/transcription.d.ts.map +1 -0
- package/dist/src/index.d.ts +22 -0
- package/dist/src/index.d.ts.map +1 -0
- package/dist/src/plugin.d.ts +19 -0
- package/dist/src/plugin.d.ts.map +1 -0
- package/dist/src/utils/error-handler.d.ts +25 -0
- package/dist/src/utils/error-handler.d.ts.map +1 -0
- package/dist/tsconfig.build.tsbuildinfo +1 -0
- package/package.json +132 -0
package/README.md
ADDED
|
@@ -0,0 +1,94 @@
|
|
|
1
|
+
# @ghostspeak/plugin-gateway-ghost
|
|
2
|
+
|
|
3
|
+
Unified Vercel AI Gateway plugin for ElizaOS - **one API key** for all model types.
|
|
4
|
+
|
|
5
|
+
## Features
|
|
6
|
+
|
|
7
|
+
- **Single API Key**: One `AI_GATEWAY_API_KEY` for all model operations
|
|
8
|
+
- **January 2026 Models**: Claude Haiku/Opus 4.5, FLUX 2 Pro, Grok 2 Vision
|
|
9
|
+
- **Full Model Coverage**: Text, embeddings, images, vision, speech, transcription
|
|
10
|
+
- **Streaming Support**: Full async streaming for text generation
|
|
11
|
+
- **Local Tokenization**: Uses js-tiktoken, no API calls needed
|
|
12
|
+
|
|
13
|
+
## Supported Model Types
|
|
14
|
+
|
|
15
|
+
| ModelType | Default Model | Description |
|
|
16
|
+
|-----------|---------------|-------------|
|
|
17
|
+
| TEXT_SMALL | `anthropic/claude-haiku-4.5` | Fast, efficient text |
|
|
18
|
+
| TEXT_LARGE | `anthropic/claude-opus-4.5` | Best quality reasoning |
|
|
19
|
+
| TEXT_EMBEDDING | `openai/text-embedding-3-large` | 3072-dim embeddings |
|
|
20
|
+
| OBJECT_SMALL | `anthropic/claude-haiku-4.5` | Structured JSON output |
|
|
21
|
+
| OBJECT_LARGE | `anthropic/claude-opus-4.5` | Complex structured output |
|
|
22
|
+
| IMAGE | `bfl/flux-2-pro` | Image generation (FLUX) |
|
|
23
|
+
| IMAGE_DESCRIPTION | `xai/grok-2-vision` | Vision/multimodal |
|
|
24
|
+
| TRANSCRIPTION | `openai/whisper-1` | Speech-to-text |
|
|
25
|
+
| TEXT_TO_SPEECH | `openai/tts-1-hd` | Text-to-speech (HD) |
|
|
26
|
+
| TEXT_TOKENIZER_* | (local) | Token encode/decode |
|
|
27
|
+
|
|
28
|
+
## Installation
|
|
29
|
+
|
|
30
|
+
```bash
|
|
31
|
+
npm install @ghostspeak/plugin-gateway-ghost
|
|
32
|
+
# or
|
|
33
|
+
bun add @ghostspeak/plugin-gateway-ghost
|
|
34
|
+
```
|
|
35
|
+
|
|
36
|
+
## Configuration
|
|
37
|
+
|
|
38
|
+
### Environment Variables
|
|
39
|
+
|
|
40
|
+
```bash
|
|
41
|
+
# Required
|
|
42
|
+
AI_GATEWAY_API_KEY=your_vercel_ai_gateway_key
|
|
43
|
+
|
|
44
|
+
# Optional (with defaults)
|
|
45
|
+
AI_GATEWAY_BASE_URL=https://ai-gateway.vercel.sh/v1
|
|
46
|
+
AI_GATEWAY_TEXT_SMALL_MODEL=anthropic/claude-haiku-4.5
|
|
47
|
+
AI_GATEWAY_TEXT_LARGE_MODEL=anthropic/claude-opus-4.5
|
|
48
|
+
AI_GATEWAY_EMBEDDING_MODEL=openai/text-embedding-3-large
|
|
49
|
+
AI_GATEWAY_IMAGE_MODEL=bfl/flux-2-pro
|
|
50
|
+
AI_GATEWAY_VISION_MODEL=xai/grok-2-vision
|
|
51
|
+
AI_GATEWAY_TRANSCRIPTION_MODEL=openai/whisper-1
|
|
52
|
+
AI_GATEWAY_TTS_MODEL=openai/tts-1-hd
|
|
53
|
+
AI_GATEWAY_TTS_VOICE=alloy
|
|
54
|
+
```
|
|
55
|
+
|
|
56
|
+
### Character Configuration
|
|
57
|
+
|
|
58
|
+
```json
|
|
59
|
+
{
|
|
60
|
+
"name": "YourAgent",
|
|
61
|
+
"plugins": ["@ghostspeak/plugin-gateway-ghost"],
|
|
62
|
+
"bio": "Your agent bio..."
|
|
63
|
+
}
|
|
64
|
+
```
|
|
65
|
+
|
|
66
|
+
## Usage
|
|
67
|
+
|
|
68
|
+
The plugin automatically registers model handlers when loaded. ElizaOS will use these handlers for all model operations:
|
|
69
|
+
|
|
70
|
+
```typescript
|
|
71
|
+
// The plugin registers handlers for:
|
|
72
|
+
// - runtime.generateText() -> TEXT_SMALL/TEXT_LARGE
|
|
73
|
+
// - runtime.embed() -> TEXT_EMBEDDING
|
|
74
|
+
// - runtime.generateObject() -> OBJECT_SMALL/OBJECT_LARGE
|
|
75
|
+
// - runtime.generateImage() -> IMAGE
|
|
76
|
+
// - etc.
|
|
77
|
+
```
|
|
78
|
+
|
|
79
|
+
## Development
|
|
80
|
+
|
|
81
|
+
```bash
|
|
82
|
+
# Install dependencies
|
|
83
|
+
bun install
|
|
84
|
+
|
|
85
|
+
# Build
|
|
86
|
+
bun run build
|
|
87
|
+
|
|
88
|
+
# Test (requires AI_GATEWAY_API_KEY)
|
|
89
|
+
AI_GATEWAY_API_KEY=your_key bun test
|
|
90
|
+
```
|
|
91
|
+
|
|
92
|
+
## License
|
|
93
|
+
|
|
94
|
+
MIT
|