yumoqing
4e65ff8fe4
Replaces wrong-direction llm_api.py (which served OpenAI endpoints) with llm_client.py -- a client that calls supplier LLM APIs. New module: llm_client.py - llm_chat(messages, model, temperature, ...) -> OpenAI response dict - llm_chat_stream(messages, ...) -> async generator of SSE chunks - llm_simple(prompt, system) -> plain text response - llm_list_models() -> list available models from provider - llm_get_config() -> show current config (key masked) - Supports provider presets: openai, dashscope, deepseek, siliconflow - Retry with exponential backoff (3 attempts) - 429 rate limit handling with Retry-After - Structured logging via appPublic.log Model changes (harnessed_agent_config): - Add llm_provider (preset name: dashscope/openai/deepseek/siliconflow) - Add top_p field - llm_service_url defaults to DashScope compatible endpoint Other: - Remove wrong-direction /v1/ endpoints - Fix pyproject.toml deps: only sqlor + bricks_for_python - Update init/data.json seed data with LLM config fields - Update CRUD view with llm_provider dropdown
173 lines
5.0 KiB
JSON
173 lines
5.0 KiB
JSON
{
|
|
"summary": [
|
|
{
|
|
"name": "harnessed_agent_config",
|
|
"title": "Hermes Agent module configuration settings",
|
|
"primary": "id"
|
|
}
|
|
],
|
|
"fields": [
|
|
{
|
|
"name": "id",
|
|
"title": "Unique configuration identifier",
|
|
"type": "str",
|
|
"length": 32,
|
|
"nullable": "no"
|
|
},
|
|
{
|
|
"name": "user_id",
|
|
"title": "User ID for multi-user isolation",
|
|
"type": "str",
|
|
"length": 32,
|
|
"nullable": "no"
|
|
},
|
|
{
|
|
"name": "work_dir",
|
|
"title": "Working directory path",
|
|
"type": "str",
|
|
"length": 255,
|
|
"nullable": "no",
|
|
"default": "./hermes_work"
|
|
},
|
|
{
|
|
"name": "skills_path",
|
|
"title": "Skills directory path",
|
|
"type": "str",
|
|
"length": 255,
|
|
"nullable": "no",
|
|
"default": "~/.hermes/skills"
|
|
},
|
|
{
|
|
"name": "max_memory_tokens",
|
|
"title": "Maximum tokens for memory context",
|
|
"type": "int",
|
|
"nullable": "no",
|
|
"default": "2000"
|
|
},
|
|
{
|
|
"name": "default_priority",
|
|
"title": "Default priority for new memories (0-100)",
|
|
"type": "int",
|
|
"nullable": "no",
|
|
"default": "50"
|
|
},
|
|
{
|
|
"name": "high_priority_threshold",
|
|
"title": "Threshold for high priority memories",
|
|
"type": "int",
|
|
"nullable": "no",
|
|
"default": "70"
|
|
},
|
|
{
|
|
"name": "low_priority_threshold",
|
|
"title": "Threshold for low priority memories",
|
|
"type": "int",
|
|
"nullable": "no",
|
|
"default": "30"
|
|
},
|
|
{
|
|
"name": "auto_cleanup_enabled",
|
|
"title": "Enable automatic memory cleanup",
|
|
"type": "str",
|
|
"length": "1",
|
|
"nullable": "no",
|
|
"default": "1"
|
|
},
|
|
{
|
|
"name": "min_retention_days",
|
|
"title": "Minimum days to retain memories",
|
|
"type": "int",
|
|
"nullable": "no",
|
|
"default": "30"
|
|
},
|
|
{
|
|
"name": "default_model",
|
|
"title": "Default LLM model for agent tasks",
|
|
"type": "str",
|
|
"length": 64,
|
|
"nullable": "yes",
|
|
"default": "qwen3-max"
|
|
},
|
|
{
|
|
"name": "default_temperature",
|
|
"title": "Default temperature for LLM calls",
|
|
"type": "float",
|
|
"length": 5,
|
|
"dec": 2,
|
|
"nullable": "no",
|
|
"default": "0.7"
|
|
},
|
|
{
|
|
"name": "top_p",
|
|
"title": "Default top_p for LLM calls",
|
|
"type": "float",
|
|
"length": 5,
|
|
"dec": 2,
|
|
"nullable": "no",
|
|
"default": "1.00"
|
|
},
|
|
{
|
|
"name": "enable_streaming",
|
|
"title": "Enable streaming response for LLM calls",
|
|
"type": "str",
|
|
"length": "1",
|
|
"nullable": "no",
|
|
"default": "1"
|
|
},
|
|
{
|
|
"name": "llm_provider",
|
|
"title": "LLM provider preset name",
|
|
"type": "str",
|
|
"length": 32,
|
|
"nullable": "yes",
|
|
"default": "dashscope",
|
|
"comments": "Provider preset: openai, dashscope, deepseek, siliconflow, or custom (empty)"
|
|
},
|
|
{
|
|
"name": "llm_service_url",
|
|
"title": "LLM service base URL (OpenAI-compatible endpoint)",
|
|
"type": "str",
|
|
"length": 255,
|
|
"nullable": "yes",
|
|
"default": "https://dashscope.aliyuncs.com/compatible-mode/v1",
|
|
"comments": "Base URL for the LLM provider API, e.g. https://api.openai.com/v1"
|
|
},
|
|
{
|
|
"name": "llm_api_key",
|
|
"title": "LLM service API key",
|
|
"type": "str",
|
|
"length": 255,
|
|
"nullable": "yes",
|
|
"default": "",
|
|
"comments": "API key for LLM service authentication (Bearer token)"
|
|
},
|
|
{
|
|
"name": "available_models",
|
|
"title": "Available LLM models (JSON array)",
|
|
"type": "text",
|
|
"nullable": "yes",
|
|
"default": "",
|
|
"comments": "JSON array of model IDs, e.g. [\"qwen3-max\", \"qwen3-plus\"]"
|
|
},
|
|
{
|
|
"name": "created_at",
|
|
"title": "Creation timestamp",
|
|
"type": "timestamp",
|
|
"nullable": "no"
|
|
},
|
|
{
|
|
"name": "updated_at",
|
|
"title": "Last update timestamp",
|
|
"type": "timestamp",
|
|
"nullable": "no"
|
|
}
|
|
],
|
|
"indexes": [
|
|
{
|
|
"name": "idx_user_agent_config",
|
|
"idxtype": "index",
|
|
"idxfields": ["user_id"]
|
|
}
|
|
],
|
|
"codes": []
|
|
} |