Compose Tools

Configuration File: compose_tools.json Tool Type: Local Tools Count: 8

This page contains all tools defined in the compose_tools.json configuration file.

Available Tools

BiomarkerDiscoveryWorkflow (Type: ComposeTool)

Discover and validate biomarkers for a specific disease condition using literature analysis, expr…

BiomarkerDiscoveryWorkflow tool specification

Tool Information:

• Name: BiomarkerDiscoveryWorkflow

• Type: ComposeTool

• Description: Discover and validate biomarkers for a specific disease condition using literature analysis, expression data, pathway enrichment, and clinical validation.

Parameters:

• disease_condition (string) (required) The disease condition to discover biomarkers for (e.g., ‘breast cancer’, ‘Alzheimer’s disease’)

• sample_type (string) (optional) The type of sample to analyze (e.g., ‘blood’, ‘tissue’, ‘plasma’)

Example Usage:

query = {
    "name": "BiomarkerDiscoveryWorkflow",
    "arguments": {
        "disease_condition": "example_value"
    }
}
result = tu.run(query)

ComprehensiveDrugDiscoveryPipeline (Type: ComposeTool)

Complete end-to-end drug discovery workflow from disease to optimized candidates. Identifies targ…

ComprehensiveDrugDiscoveryPipeline tool specification

Tool Information:

• Name: ComprehensiveDrugDiscoveryPipeline

• Type: ComposeTool

• Description: Complete end-to-end drug discovery workflow from disease to optimized candidates. Identifies targets, discovers lead compounds, screens for ADMET properties, assesses safety, and validates with literature.

Parameters:

• disease_efo_id (string) (required) The EFO ID of the disease for drug discovery (e.g., ‘EFO_0001074’ for Alzheimer’s disease)

Example Usage:

query = {
    "name": "ComprehensiveDrugDiscoveryPipeline",
    "arguments": {
        "disease_efo_id": "example_value"
    }
}
result = tu.run(query)

DrugSafetyAnalyzer (Type: ComposeTool)

Comprehensive drug safety analysis combining adverse event data, literature review, and molecular…

DrugSafetyAnalyzer tool specification

Tool Information:

• Name: DrugSafetyAnalyzer

• Type: ComposeTool

• Description: Comprehensive drug safety analysis combining adverse event data, literature review, and molecular information

Parameters:

• drug_name (string) (required) Name of the drug to analyze

• patient_sex (string) (optional) Filter by patient sex (optional)

• serious_events_only (boolean) (optional) Focus only on serious adverse events

Example Usage:

query = {
    "name": "DrugSafetyAnalyzer",
    "arguments": {
        "drug_name": "example_value"
    }
}
result = tu.run(query)

LiteratureSearchTool (Type: ComposeTool)

Comprehensive literature search and summary tool that searches multiple databases (EuropePMC, Ope…

LiteratureSearchTool tool specification

Tool Information:

• Name: LiteratureSearchTool

• Type: ComposeTool

• Description: Comprehensive literature search and summary tool that searches multiple databases (EuropePMC, OpenAlex, PubTator) and generates AI-powered summaries of research findings

Parameters:

• research_topic (string) (required) The research topic or query to search for in the literature

Example Usage:

query = {
    "name": "LiteratureSearchTool",
    "arguments": {
        "research_topic": "example_value"
    }
}
result = tu.run(query)

ToolDescriptionOptimizer (Type: ComposeTool)

Optimizes a tool’s description and parameter descriptions by generating test cases, executing the…

ToolDescriptionOptimizer tool specification

Tool Information:

• Name: ToolDescriptionOptimizer

• Type: ComposeTool

• Description: Optimizes a tool’s description and parameter descriptions by generating test cases, executing them, analyzing the results, and suggesting improved descriptions for both the tool and its arguments. Optionally saves a comprehensive optimization report to a file without overwriting the original.

Parameters:

• tool_config (object) (required) The full configuration of the tool to optimize.

• save_to_file (boolean) (optional) If true, save the optimized description to a file (do not overwrite the original).

• output_file (string) (optional) Optional file path to save the optimized description. If not provided, use ‘<tool_name>_optimized_description.txt’.

• max_iterations (integer) (optional) Maximum number of optimization rounds to perform.

• satisfaction_threshold (number) (optional) Quality score threshold (1-10) to consider optimization satisfactory.

Example Usage:

query = {
    "name": "ToolDescriptionOptimizer",
    "arguments": {
        "tool_config": "example_value"
    }
}
result = tu.run(query)

ToolDiscover (Type: ComposeTool)

Generates new ToolUniverse-compliant tools based on short descriptions through an intelligent dis…

ToolDiscover tool specification

Tool Information:

• Name: ToolDiscover

• Type: ComposeTool

• Description: Generates new ToolUniverse-compliant tools based on short descriptions through an intelligent discovery and refinement process. Automatically determines the optimal tool type and category, discovers similar existing tools, generates initial specifications, and iteratively refines the tool configuration using agentic optimization tools until it meets quality standards.

Parameters:

• tool_description (string) (required) Short description of the desired tool functionality and purpose. Tool Discover will automatically analyze this to determine the optimal tool type (PackageTool, RESTTool, XMLTool, or AgenticTool) and appropriate category.

• max_iterations (integer) (optional) Maximum number of refinement iterations to perform.

• save_to_file (boolean) (optional) Whether to save the generated tool configuration and report to a file.

• output_file (string) (optional) Optional file path to save the generated tool. If not provided, uses auto-generated filename.

Example Usage:

query = {
    "name": "ToolDiscover",
    "arguments": {
        "tool_description": "example_value"
    }
}
result = tu.run(query)

ToolGraphGenerationPipeline (Type: ComposeTool)

Generates a directed tool relationship graph among provided tool configs using ToolRelationshipDe…

ToolGraphGenerationPipeline tool specification

Tool Information:

• Name: ToolGraphGenerationPipeline

• Type: ComposeTool

• Description: Generates a directed tool relationship graph among provided tool configs using ToolRelationshipDetector to infer data-flow compatibility.

Parameters:

• tool_configs (array) (required) List of tool configuration objects

• max_tools (integer) (optional) Optional max number of tools to process (debug)

• output_path (string) (optional) Path for output graph JSON

• save_intermediate_every (integer) (optional) Checkpoint every N processed pairs

Example Usage:

query = {
    "name": "ToolGraphGenerationPipeline",
    "arguments": {
        "tool_configs": ["item1", "item2"]
    }
}
result = tu.run(query)

ToolMetadataGenerationPipeline (Type: ComposeTool)

Generates standardized metadata for a batch of ToolUniverse tool configurations by calling ToolMe…

ToolMetadataGenerationPipeline tool specification

Tool Information:

• Name: ToolMetadataGenerationPipeline

• Type: ComposeTool

• Description: Generates standardized metadata for a batch of ToolUniverse tool configurations by calling ToolMetadataGenerator, LabelGenerator, and ToolMetadataStandardizer for sources and tags.

Parameters:

• tool_configs (array) (required) List of raw tool configuration JSON objects to extract and standardize metadata for

• tool_type_mappings (object) (optional) Mapping of simplified toolType (keys) to lists of tool ‘type’ values belonging to each simplified category (e.g., {‘Databases’: [‘XMLTool’]})

• add_existing_tooluniverse_labels (boolean) (optional) Whether to include labels from existing ToolUniverse tools when labeling the metadata configs of the new tools. It is strongly recommended that this is set to true to minimize the number of new labels created and the possibility of redundant labels.

• max_new_tooluniverse_labels (integer) (optional) The maximum number of new ToolUniverse labels to use in the metadata configs of the new tools. The existing ToolUniverse labels will be used first, and then new labels will be created as needed up to this limit. If the limit is reached, the least relevant new labels will be discarded. Please try to use as few new labels as possible to avoid excessive labels.

Example Usage:

query = {
    "name": "ToolMetadataGenerationPipeline",
    "arguments": {
        "tool_configs": ["item1", "item2"]
    }
}
result = tu.run(query)

Navigation

• Available Tools - Back to Tools Overview

• Tool Loading Tutorial - Loading Local Tools