feat(agent): implement LLM call timeout and transient error handling
- Introduced a configurable LLM call timeout with a default value, allowing for better control over LLM call durations. - Added logic to retry transient errors (e.g., timeouts, rate limits) up to a specified maximum number of retries, improving robustness in error handling. - Implemented parameter casting for tool arguments to ensure correct types are used, addressing common LLM quirks. - Enhanced tool execution error messages with guidance for retrying with different approaches. - Added validation for configuration values to prevent runtime errors.
This commit is contained in:
@@ -379,9 +379,58 @@ func LoadConfig() (*Config, error) {
|
||||
resolveBuiltinAgentPromptIDs(cfg.PromptTemplates)
|
||||
}
|
||||
|
||||
// Validate configuration values
|
||||
if err := ValidateConfig(&cfg); err != nil {
|
||||
return nil, err
|
||||
}
|
||||
|
||||
return &cfg, nil
|
||||
}
|
||||
|
||||
// ValidateConfig performs basic validation of the loaded configuration.
|
||||
// It checks for obviously invalid or missing values that would cause runtime failures.
|
||||
func ValidateConfig(cfg *Config) error {
|
||||
var errs []string
|
||||
|
||||
if cfg.Conversation != nil {
|
||||
if cfg.Conversation.EmbeddingTopK < 0 {
|
||||
errs = append(errs, "conversation.embedding_top_k must be >= 0")
|
||||
}
|
||||
if cfg.Conversation.RerankTopK < 0 {
|
||||
errs = append(errs, "conversation.rerank_top_k must be >= 0")
|
||||
}
|
||||
if cfg.Conversation.VectorThreshold < 0 || cfg.Conversation.VectorThreshold > 1 {
|
||||
errs = append(errs, "conversation.vector_threshold must be between 0 and 1")
|
||||
}
|
||||
if cfg.Conversation.RerankThreshold < 0 || cfg.Conversation.RerankThreshold > 1 {
|
||||
errs = append(errs, "conversation.rerank_threshold must be between 0 and 1")
|
||||
}
|
||||
}
|
||||
|
||||
if cfg.KnowledgeBase != nil {
|
||||
if cfg.KnowledgeBase.ChunkSize <= 0 {
|
||||
errs = append(errs, "knowledge_base.chunk_size must be > 0")
|
||||
}
|
||||
if cfg.KnowledgeBase.ChunkOverlap < 0 {
|
||||
errs = append(errs, "knowledge_base.chunk_overlap must be >= 0")
|
||||
}
|
||||
if cfg.KnowledgeBase.ChunkOverlap >= cfg.KnowledgeBase.ChunkSize {
|
||||
errs = append(errs, "knowledge_base.chunk_overlap must be less than chunk_size")
|
||||
}
|
||||
}
|
||||
|
||||
if cfg.Server != nil {
|
||||
if cfg.Server.Port <= 0 || cfg.Server.Port > 65535 {
|
||||
errs = append(errs, "server.port must be between 1 and 65535")
|
||||
}
|
||||
}
|
||||
|
||||
if len(errs) > 0 {
|
||||
return fmt.Errorf("config validation errors: %s", strings.Join(errs, "; "))
|
||||
}
|
||||
return nil
|
||||
}
|
||||
|
||||
// backfillConversationDefaults resolves prompt template ID references
|
||||
// into actual prompt text content. Only xxx_id fields are used;
|
||||
// no fallback to default templates.
|
||||
|
||||
Reference in New Issue
Block a user