ALTK Agent
The Agent Lifecycle Toolkit (ALTK) is a set of advanced capabilities designed to enhance agent behavior, reliability, and efficiency within Robility Flow. It provides intelligent controls and processing mechanisms that improve how agents interact with tools, handle responses, and execute workflows.
Key Features
a. Pre-tool Validation
Validates tool calls before execution to ensure they are appropriate and correctly structured. This is powered by the SPARC reflection component and helps prevent agents from executing invalid or unintended actions.
b. Post-tool JSON Processing
Efficiently processes large JSON responses from tools by dynamically generating Python code to extract only the relevant data. This reduces context size and improves performance when handling APIs that return extensive datasets.
Instead of passing the entire JSON response, the component outputs a concise Message containing only the necessary information, which is then forwarded to the next component in the flow.
Note
1. ALTK includes its own set of features referred to as components, which are different from standard Robility Flow components. All ALTK components are accessed and utilized through the ALTK Agent component in Robility Flow.
2. The ALTK Agent acts as an interface to implement and manage these capabilities, allowing individual features to be enabled or disabled based on specific use cases.
For more details, refer to the Agent Lifecycle Toolkit documentation.
ALTK Agent parameters
Some parameters are hidden by default in the visual editor. You can modify all component parameters through the component inspection panel that appears when you select a component.
| Name | Type | Description |
|---|---|---|
| agent_llm | Dropdown | Input parameter. The model provider the agent uses to generate responses. |
| enable_tool_validation | Boolean | Input parameter. If enabled, tool calls are validated using SPARC before execution to check for appropriateness and correctness. Default: true. |
| enable_post_tool_reflection | Boolean | Input parameter. If enabled, tool outputs are automatically processed through JSON processing when the output is JSON and exceeds the size threshold. Default: true. |
| response_processing_size_threshold | Integer | Input parameter. Tool output is post-processed only if the response length exceeds this character threshold. Default: 100. Advanced parameter. |
| tools | List[Tool] | Input parameter. The list of tools available to the agent. |
| system_prompt | String | Input parameter. The system prompt to provide context to the agent. |
| input_value | String | Input parameter. The user's input to the agent. |
| memory | Memory | Input parameter. The memory for the agent to use for context persistence. |
| max_iterations | Integer | Input parameter. The maximum number of iterations to allow the agent to execute. |
| verbose | Boolean | Input parameter. This determines whether to print out the agent's intermediate steps. |
| handle_parsing_errors | Boolean | Input parameter. This determines whether to handle parsing errors in the agent. |
