feat: golem15.golem plugin with the OpenAI-compatible adapter and AIService.Send
- Prompt, Payload and AIResponse value objects ported from Golem15.Golem - OpenAIAdapter: chat completions mapping with PHP's response_format, max_completion_tokens and reasoning_effort rules - AIService.Send through fetchguard (TrustedMode for operator models, PublicOnlyMode otherwise) with PHP's failure messages - config defaults for the SSRF allowlist and the AI timeout, lang, README
This commit is contained in:
29
lang/en/lang.yaml
Normal file
29
lang/en/lang.yaml
Normal file
@@ -0,0 +1,29 @@
|
||||
plugin:
|
||||
name: Golem
|
||||
description: Clean, extendable AI integration
|
||||
permissions:
|
||||
access_settings: Manage AI settings
|
||||
settings:
|
||||
menu_label: AI
|
||||
menu_description: Configure AI models and settings
|
||||
adapter: Adapter
|
||||
adapter_comment: OpenAI-compatible works for most providers (Perplexity, Groq, Together, etc.)
|
||||
is_enabled: Enabled
|
||||
has_files_endpoint: Has files endpoint
|
||||
max_completion_tokens: Max Completion Tokens
|
||||
max_completion_tokens_comment: "Maximum tokens in AI response (default: 4096)"
|
||||
models: AI Models
|
||||
add_model: Add Model
|
||||
model_name: Name
|
||||
api_key: API Key
|
||||
api_key_comment: Write-only. Leave it unchanged to keep the stored key.
|
||||
base_url: Base URL
|
||||
base_url_comment: OpenAI-compatible endpoint
|
||||
model_id: Model ID
|
||||
model_id_comment: e.g., gpt-4o, gpt-4o-mini, claude-3-5-sonnet
|
||||
system_prompt: System Prompt
|
||||
system_prompt_comment: Default system prompt for this model. Used when no explicit system prompt is provided by the caller.
|
||||
is_default: Default Model
|
||||
generates_images: Generates Images
|
||||
accepts_images: Accepts Images (Vision)
|
||||
sort_order: Order
|
||||
Reference in New Issue
Block a user