Update AI Agent Configurations

Used to update the AI-specific configurations for an existing agent, such as model parameters, prompt templates, RAG settings, guardrails, tools, and memory configurations.

API Endpoint

Property	Value
Request Method	POST
Request URL	https://api.seliseblocks.com/agents/update-ai-configurations

Request

Request Example

curl -X POST 'https://api.seliseblocks.com/agents/update-ai-configurations' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
    "enable_rag": true,
    "rag_config": {
      "TopK": 5,
      "ScoreThreshold": 0.3,
      "EmbeddingModel": "text-embedding-ada-002",
      "EmbeddingModelProvider": "azure",
      "ChunkingStrategy": "recursive",
      "EnableQueryEnhancement": true,
      "EnableRerank": true,
      "DenseWeight": 0.9,
      "KeywordWeight": 0.1
    },
    "llm_config": {
      "Provider": "openai",
      "ModelName": "gpt-4",
      "IsBlocksProvider": true,
      "ModelId": "gpt-4-turbo",
      "Temperature": 0.7,
      "MaxTokens": 2000,
      "ModelType": "chat",
      "BasePrompt": "You are a helpful assistant",
      "BasePromptToken": 50
    },
    "enable_guardrails": true,
    "guardrail": {
      "EnablePiiDetection": false,
      "EnableInjectionDetection": false,
      "EnableKeywordCheck": false,
      "EnableCustomRules": false,
      "EnablePreValidation": true,
      "EnablePostValidation": true,
      "BannedKeywords": ["spam", "inappropriate"],
      "BannedPatterns": [".*badpattern.*"],
      "CustomRules": [
        {
          "Name": "Rule1",
          "Pattern": ".*",
          "Severity": "medium",
          "AutoBlock": false,
          "Message": "Custom rule triggered",
          "EnablePre": true,
          "EnablePost": true
        }
      ],
      "InputRiskThreshold": 0.7,
      "OutputRiskThreshold": 0.8
    },
    "enable_tools": true,
    "tool_ids": ["tool_id_1", "tool_id_2"],
    "memory_config": {
      "LastNConversations": 5,
      "EnableSummary": false,
      "EnableQuestionSuggestions": false
    },
    "response_type": "text",
    "response_format": "markdown",
    "project_key": "YOUR_PROJECT_KEY",
    "agent_id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890"
  }'

Request Headers

Field	Type	Required	Description
accept	string	Yes	Accepted response format. Use application/json
Content-Type	application/json	Yes	Data type, must be application/json.

Request Body

Request Body Schema

{
  "enable_rag": true,
  "rag_config": {
    "TopK": 5,
    "ScoreThreshold": 0.3,
    "EmbeddingModel": "text-embedding-ada-002",
    "EmbeddingModelProvider": "azure",
    "ChunkingStrategy": "recursive",
    "EnableQueryEnhancement": true,
    "EnableRerank": true,
    "DenseWeight": 0.9,
    "KeywordWeight": 0.1
  },
  "llm_config": {
    "Provider": "",
    "ModelName": "",
    "IsBlocksProvider": true,
    "ModelId": "string",
    "Temperature": 0.7,
    "MaxTokens": 0,
    "ModelType": "chat",
    "BasePrompt": "string",
    "BasePromptToken": 0
  },
  "enable_guardrails": true,
  "guardrail": {
    "EnablePiiDetection": false,
    "EnableInjectionDetection": false,
    "EnableKeywordCheck": false,
    "EnableCustomRules": false,
    "EnablePreValidation": true,
    "EnablePostValidation": true,
    "BannedKeywords": ["string"],
    "BannedPatterns": ["string"],
    "CustomRules": [
      {
        "Name": "",
        "Pattern": "",
        "Severity": "medium",
        "AutoBlock": false,
        "Message": "",
        "EnablePre": true,
        "EnablePost": true
      }
    ],
    "InputRiskThreshold": 0.7,
    "OutputRiskThreshold": 0.8
  },
  "enable_tools": true,
  "tool_ids": ["string"],
  "memory_config": {
    "LastNConversations": 5,
    "EnableSummary": false,
    "EnableQuestionSuggestions": false
  },
  "response_type": "text",
  "response_format": "string",
  "project_key": "string",
  "agent_id": "string"
}

Request Body Parameters

Field	Type	Required	Description
enable_rag	boolean	No	Enable Retrieval-Augmented Generation (RAG) for the agent.
rag_config	object	No	Configuration settings for RAG functionality.
llm_config	object	Yes	Large Language Model configuration settings.
enable_guardrails	boolean	No	Enable safety guardrails for content moderation.
guardrail	object	No	Guardrail configuration settings.
enable_tools	boolean	No	Enable external tools/integrations for the agent.
tool_ids	array	No	Array of tool identifiers to enable for the agent.
memory_config	object	No	Configuration for conversation memory and context.
response_type	string	No	Type of response output (e.g., text, json).
response_format	string	No	Format specification for the response.
project_key	string	Yes	The project key for your project.
agent_id	string	Yes	Unique identifier of the agent to be updated.

RAG Configuration Parameters

Field	Type	Description
TopK	integer	Number of top results to retrieve.
ScoreThreshold	number	Minimum similarity score threshold for retrieved documents.
EmbeddingModel	string	Name of the embedding model to use.
EmbeddingModelProvider	string	Provider of the embedding model (e.g., azure, openai).
ChunkingStrategy	string	Strategy for splitting documents into chunks.
EnableQueryEnhancement	boolean	Enable query enhancement for better retrieval.
EnableRerank	boolean	Enable reranking of retrieved results.
DenseWeight	number	Weight for dense retrieval (0.0-1.0).
KeywordWeight	number	Weight for keyword-based retrieval (0.0-1.0).

LLM Configuration Parameters

Field	Type	Description
Provider	string	LLM provider name (e.g., openai, azure).
ModelName	string	Name of the language model.
IsBlocksProvider	boolean	Whether using Blocks-managed provider.
ModelId	string	Unique identifier for the model.
Temperature	number	Controls randomness in responses (0.0-1.0).
MaxTokens	integer	Maximum number of tokens in the response.
ModelType	string	Type of model (e.g., chat, completion).
BasePrompt	string	System prompt or base instructions for the model.
BasePromptToken	integer	Token count for the base prompt.

Guardrail Configuration Parameters

Field	Type	Description
EnablePiiDetection	boolean	Enable detection of Personally Identifiable Information.
EnableInjectionDetection	boolean	Enable detection of prompt injection attempts.
EnableKeywordCheck	boolean	Enable checking for banned keywords.
EnableCustomRules	boolean	Enable custom validation rules.
EnablePreValidation	boolean	Enable validation before processing.
EnablePostValidation	boolean	Enable validation after processing.
BannedKeywords	array	List of keywords to block.
BannedPatterns	array	List of regex patterns to block.
CustomRules	array	Array of custom rule objects.
InputRiskThreshold	number	Risk threshold for input validation (0.0-1.0).
OutputRiskThreshold	number	Risk threshold for output validation (0.0-1.0).

Custom Rule Parameters

Field	Type	Description
Name	string	Name of the custom rule.
Pattern	string	Regex pattern to match.
Severity	string	Severity level (e.g., low, medium, high).
AutoBlock	boolean	Automatically block content matching this rule.
Message	string	Message to display when rule is triggered.
EnablePre	boolean	Enable rule for pre-validation.
EnablePost	boolean	Enable rule for post-validation.

Memory Configuration Parameters

Field	Type	Description
LastNConversations	integer	Number of recent conversations to keep in memory.
EnableSummary	boolean	Enable conversation summarization.
EnableQuestionSuggestions	boolean	Enable AI-generated follow-up question suggestions.

Response

Success Response (200 OK)

Returns an object containing the update status and details.

{
  "is_success": true,
  "item_id": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
  "detail": "AI configurations updated successfully",
  "error": {}
}

Response Fields

Field	Type	Description
is_success	boolean	Indicates whether the configuration update was successful.
item_id	string	Unique identifier for the updated agent.
detail	string	Success or failure message with additional context.
error	object	Error details if the operation failed (empty if successful).

Error Response (422 Unprocessable Entity)

Returns validation error details when the request body is invalid.

{
  "detail": [
    {
      "loc": [
        "body",
        "agent_id"
      ],
      "msg": "field required",
      "type": "value_error.missing"
    }
  ]
}

Error Response Fields

Field	Type	Description
detail	array	Array of validation error objects.
loc	array	Location of the error in the request (e.g., body field).
msg	string	Human-readable error message.
type	string	Error type identifier.

Error Codes

Status Code	Description	Response Type
200	Successful Response	Success
422	Validation Error - Invalid request parameters	Unprocessable Entity