Skip to content

fix(AutoTradingAgent): support other providers #242

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
hazeone merged 2 commits into from
Oct 29, 2025
Merged

fix(AutoTradingAgent): support other providers #242

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
038de4d
opt agent card, remove additional info
hazeone Oct 29, 2025
de50e50
opt model provider fallback logic
hazeone Oct 29, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
137 changes: 14 additions & 123 deletions python/configs/agent_cards/auto_trading_agent.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,67 +8,20 @@
"push_notifications": true
},
"skills": [{
"id": "auto_trading_setup",
"name": "Setup Auto Trading",
"description": "Configure the automated trading agent with initial capital, crypto symbols, risk parameters, and AI model selection.",
"examples": [
"Setup trading with $100,000 for BTC-USD and ETH-USD using DeepSeek, Grok, Claude model",
"Configure auto trader with $50,000 capital, trade BTC-USD, ETH-USD, SOL-USD with 1.5% risk",
"Setup trading agent with default settings for Bitcoin"
],
"tags": [
"setup",
"configuration",
"initialization"
]
},
{
"id": "auto_trading_notify",
"name": "Live Trading Monitoring",
"description": "Start continuous automated trading with real-time technical analysis, trade execution notifications, and portfolio updates.",
"examples": [
"Start automated trading for BTC-USD with Claude model",
"Monitor and trade ETH-USD and SOL-USD using GPT-4 analysis",
"Begin live trading with portfolio notifications"
],
"tags": [
"trading",
"monitoring",
"real-time",
"notifications"
]
},
{
"id": "technical_analysis",
"name": "Technical Analysis",
"description": "Calculate and analyze technical indicators including MACD, RSI, EMA, and Bollinger Bands for trading decisions.",
"examples": [
"Analyze BTC-USD technical indicators",
"Show MACD and RSI signals for ETH-USD",
"Calculate Bollinger Bands for SOL-USD"
],
"tags": [
"analysis",
"indicators",
"signals"
]
},
{
"id": "position_management",
"name": "Position Management",
"description": "Manage trading positions with risk controls, automatic entry/exit, and portfolio tracking.",
"examples": [
"Show current positions and portfolio value",
"What is my P&L on open trades?",
"Display portfolio summary"
],
"tags": [
"positions",
"portfolio",
"risk-management"
]
}
],
"id": "auto_trading_setup",
"name": "Setup Auto Trading",
"description": "Configure the automated trading agent with initial capital, crypto symbols, risk parameters, and AI model selection.",
"examples": [
"Setup trading with $100,000 for BTC-USD and ETH-USD using DeepSeek, Grok, Claude model",
"Configure auto trader with $50,000 capital, trade BTC-USD, ETH-USD, SOL-USD with 1.5% risk",
"Setup trading agent with default settings for Bitcoin"
],
"tags": [
"setup",
"configuration",
"initialization"
]
}],
"enabled": true,
"provider": {
"organization": "ValueCell",
Expand All @@ -85,74 +38,12 @@
"ai-trading",
"real-time"
],
"supported_crypto_symbols": [
"BTC-USD",
"ETH-USD",
"USDT-USD",
"BNB-USD",
"XRP-USD",
"SOL-USD",
"USDC-USD",
"TRX-USD",
"DOGE-USD",
"ADA-USD",
"MATIC-USD",
"WBTC-USD",
"LINK-USD",
"XLM-USD",
"BCH-USD",
"SUI-USD",
"HBAR-USD",
"AVAX-USD",
"TON-USD",
"LTC-USD",
"MNT-USD",
"NEAR-USD",
"FLOW-USD",
"OKB-USD",
"APE-USD",
"ATOM-USD",
"QNT-USD",
"BSV-USD",
"EOS-USD",
"NEO-USD",
"DYDX-USD",
"CRV-USD",
"KLAY-USD",
"MINA-USD",
"CHZ-USD",
"AXS-USD",
"FTM-USD",
"IMX-USD",
"ZEC-USD",
"ALGO-USD",
"XDC-USD",
"ICP-USD",
"BTG-USD",
"BLZ-USD",
"ENJ-USD",
"KSM-USD",
"VGX-USD",
"RUNE-USD",
"BAT-USD",
"THETA-USD",
"QTUM-USD"
],
"technical_indicators": [
"MACD (12/26/9)",
"RSI (14)",
"EMA (12/26/50)",
"Bollinger Bands (20, 2σ)"
],
"supported_models": [
"deepseek/deepseek-v3.1-terminus",
"openai/gpt-5",
"anthropic/claude-sonnet-4.5",
"x-ai/grok-4",
"qwen/qwen-max",
"google/gemini-2.5-pro",
"meta-llama/llama-3.1-70b-instruct"
],
"trading_features": [
"Long and Short positions",
"Risk-based position sizing",
Expand Down
10 changes: 7 additions & 3 deletions python/configs/agents/auto_trading_agent.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,10 +15,14 @@ enabled: true
models:
primary:
model_id: "deepseek/deepseek-v3.1-terminus"
provider: openrouter
provider: "openrouter" # Must explicitly specify provider (not null)
# Provider-specific model mappings for fallback
provider_models:
openrouter: "anthropic/claude-haiku-4.5" # Fast and reasoning capable
google: "gemini-2.5-flash" # Fast and efficient
parameters:
temperature: 0.5
max_tokens: 2048
# temperature: 0.5
# max_tokens: 2048

# Environment Variable Overrides
# Format: ENV_VAR_NAME -> config.path
Expand Down
9 changes: 8 additions & 1 deletion python/configs/agents/research_agent.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,10 @@ models:
primary:
model_id: "google/gemini-2.5-flash"
provider: "openrouter"
# Provider-specific model mappings for fallback
provider_models:
siliconflow: "Qwen/Qwen3-235B-A22B-Thinking-2507"
google: "gemini-2.5-flash"
parameters:
# temperature: 0.7
# max_tokens: 8192
Expand All @@ -19,7 +23,10 @@ models:
# Note: If not specified, will auto-select from available providers
embedding:
model_id: "Qwen/Qwen3-Embedding-4B" # Optional, uses provider default
provider: "siliconflow" # Optional, auto-detects if not specified
provider: "siliconflow" # Must explicitly specify provider
# Provider-specific embedding model mappings for fallback
provider_models:
google: "gemini-embedding-001"
parameters:
dimensions: ${EMBEDDER_DIMENSION:2560}
# encoding_format: "float"
Expand Down
8 changes: 7 additions & 1 deletion python/configs/agents/super_agent.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,13 @@ models:
# Primary model for triage, reasoning, and query enrichment
primary:
model_id: "anthropic/claude-haiku-4.5"
provider: "openrouter" # Can be: openrouter, siliconflow, or null (uses system primary_provider)
provider: "openrouter" # Must explicitly specify provider (not null)

# Provider-specific model mappings for fallback
# Used when primary provider fails and fallback is attempted
provider_models:
siliconflow: "zai-org/GLM-4.6" # Similar capability to Claude Haiku
google: "gemini-2.5-flash" # Fast and efficient like Claude Haiku

# Environment Variable Overrides
# Format: ENV_VAR_NAME -> config.path
Expand Down
121 changes: 115 additions & 6 deletions python/valuecell/adapters/models/factory.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -296,6 +296,7 @@ def create_model(
model_id: Optional[str] = None,
provider: Optional[str] = None,
use_fallback: bool = True,
provider_models: Optional[dict] = None,
**kwargs,
):
"""
Expand All @@ -310,6 +311,8 @@ def create_model(
model_id: Specific model ID (optional, uses provider default)
provider: Provider name (optional, uses primary_provider)
use_fallback: Try fallback providers if primary fails
provider_models: Dict mapping provider names to model IDs for fallback
e.g., {"siliconflow": "deepseek-ai/DeepSeek-V3.1-Terminus"}
**kwargs: Additional arguments for model creation

Returns:
Expand All @@ -323,8 +326,16 @@ def create_model(
>>> model = factory.create_model() # Uses primary provider + default model
>>> model = factory.create_model(provider="google") # Specific provider
>>> model = factory.create_model(model_id="gpt-4", provider="openrouter")
>>> # With provider-specific models
>>> model = factory.create_model(
... model_id="anthropic/claude-3.5-sonnet",
... provider="openrouter",
... provider_models={"siliconflow": "deepseek-ai/DeepSeek-V3.1-Terminus"}
... )
"""
provider = provider or self.config_manager.primary_provider
if provider_models is None:
provider_models = {}

# Try primary provider
try:
Expand All @@ -341,9 +352,29 @@ def create_model(
continue # Skip already tried provider

try:
# Determine model ID for fallback provider
fallback_model_id = model_id

# Priority: provider_models > default_model
if fallback_provider in provider_models:
fallback_model_id = provider_models[fallback_provider]
logger.info(
f"Using provider-specific model for {fallback_provider}: {fallback_model_id}"
)
else:
# Use provider's default model
fallback_provider_config = (
self.config_manager.get_provider_config(fallback_provider)
)
if fallback_provider_config:
fallback_model_id = fallback_provider_config.default_model
logger.info(
f"Using default model for {fallback_provider}: {fallback_model_id}"
)

logger.info(f"Trying fallback provider: {fallback_provider}")
return self._create_model_internal(
model_id, fallback_provider, **kwargs
fallback_model_id, fallback_provider, **kwargs
)
except Exception as fallback_error:
logger.warning(
Expand Down Expand Up @@ -435,11 +466,44 @@ def create_model_for_agent(
f"model_id={model_config.model_id}, provider={model_config.provider}"
)

# Check if specified provider is available (has API key)
provider = model_config.provider
model_id = model_config.model_id
is_valid, error_msg = self.config_manager.validate_provider(provider)

if not is_valid:
# If configured provider is not available, use primary provider instead
fallback_provider = self.config_manager.primary_provider
logger.warning(
f"Configured provider '{provider}' for agent '{agent_name}' is not available: {error_msg}. "
f"Falling back to primary provider: {fallback_provider}"
)
provider = fallback_provider

# Update model_id for the fallback provider
# Priority: provider_models[fallback_provider] > provider's default_model
if fallback_provider in model_config.provider_models:
model_id = model_config.provider_models[fallback_provider]
logger.info(
f"Using provider-specific model for fallback: {fallback_provider} -> {model_id}"
)
else:
# Use provider's default model
provider_config = self.config_manager.get_provider_config(
fallback_provider
)
if provider_config:
model_id = provider_config.default_model
logger.info(
f"Using default model for fallback provider '{fallback_provider}': {model_id}"
)

# Create model
return self.create_model(
model_id=model_config.model_id,
provider=model_config.provider,
model_id=model_id,
provider=provider,
use_fallback=use_fallback,
provider_models=model_config.provider_models,
**merged_params,
)

Expand Down Expand Up @@ -470,9 +534,42 @@ def create_embedder_for_agent(
logger.info(
f"Creating embedder for agent '{agent_name}': model_id={emb.model_id}, provider={emb.provider}, params={merged_params}"
)

# Check if specified provider is available (has API key)
provider = emb.provider
model_id = emb.model_id
is_valid, error_msg = self.config_manager.validate_provider(provider)

if not is_valid:
# If configured provider is not available, use primary provider instead
fallback_provider = self.config_manager.primary_provider
logger.warning(
f"Configured embedding provider '{provider}' for agent '{agent_name}' is not available: {error_msg}. "
f"Falling back to primary provider: {fallback_provider}"
)
provider = fallback_provider

# Update model_id for the fallback provider
# Priority: provider_models[fallback_provider] > provider's default_embedding_model
if fallback_provider in emb.provider_models:
model_id = emb.provider_models[fallback_provider]
logger.info(
f"Using provider-specific embedding model for fallback: {fallback_provider} -> {model_id}"
)
else:
# Use provider's default embedding model
provider_config = self.config_manager.get_provider_config(
fallback_provider
)
if provider_config and provider_config.default_embedding_model:
model_id = provider_config.default_embedding_model
logger.info(
f"Using default embedding model for fallback provider '{fallback_provider}': {model_id}"
)

return self.create_embedder(
model_id=emb.model_id or None,
provider=emb.provider,
model_id=model_id or None,
provider=provider,
use_fallback=use_fallback,
**merged_params,
)
Expand All @@ -483,9 +580,21 @@ def create_embedder_for_agent(
logger.info(
f"Creating embedder for agent '{agent_name}' using primary provider: {primary.provider}"
)

# Check if primary provider is available
provider = primary.provider
is_valid, error_msg = self.config_manager.validate_provider(provider)

if not is_valid:
logger.warning(
f"Primary provider '{provider}' for agent '{agent_name}' is not available: {error_msg}. "
f"Using system primary provider: {self.config_manager.primary_provider}"
)
provider = self.config_manager.primary_provider

return self.create_embedder(
model_id=None,
provider=primary.provider,
provider=provider,
use_fallback=use_fallback,
**merged_params,
)
Expand Down
Loading