10 KiB

Raw Blame History

MALT

Specialized framework for complex language-based tasks and processing

Swarm Type: MALT

Overview

MALT (Multi-Agent Language Task) is a specialized framework optimized for complex language-based tasks, optimizing agent collaboration for sophisticated language processing operations. This architecture excels at tasks requiring deep linguistic analysis, natural language understanding, and complex text generation workflows.

Key features:

Language Optimization: Specifically designed for natural language tasks
Linguistic Collaboration: Agents work together on complex language operations
Text Processing Pipeline: Structured approach to language task workflows
Advanced NLP: Optimized for sophisticated language understanding tasks

Use Cases

Complex document analysis and processing
Multi-language translation and localization
Advanced content generation and editing
Linguistic research and analysis tasks

API Usage

Basic MALT Example

=== "Shell (curl)" bash curl -X POST "https://api.swarms.world/v1/swarm/completions" \ -H "x-api-key: $SWARMS_API_KEY" \ -H "Content-Type: application/json" \ -d '{ "name": "Legal Document Analysis MALT", "description": "Advanced linguistic analysis of legal documents using MALT framework", "swarm_type": "MALT", "task": "Perform comprehensive linguistic analysis of a complex legal contract including sentiment analysis, risk identification, clause categorization, and language complexity assessment", "agents": [ { "agent_name": "Syntactic Analyzer", "description": "Analyzes sentence structure and grammar", "system_prompt": "You are a syntactic analysis expert. Analyze sentence structure, grammatical patterns, and linguistic complexity in legal texts.", "model_name": "gpt-4o", "max_loops": 1, "temperature": 0.2 }, { "agent_name": "Semantic Analyzer", "description": "Analyzes meaning and semantic relationships", "system_prompt": "You are a semantic analysis expert. Extract meaning, identify semantic relationships, and analyze conceptual content in legal documents.", "model_name": "gpt-4o", "max_loops": 1, "temperature": 0.3 }, { "agent_name": "Pragmatic Analyzer", "description": "Analyzes context and implied meanings", "system_prompt": "You are a pragmatic analysis expert. Analyze contextual meaning, implied obligations, and pragmatic implications in legal language.", "model_name": "gpt-4o", "max_loops": 1, "temperature": 0.4 }, { "agent_name": "Discourse Analyzer", "description": "Analyzes document structure and flow", "system_prompt": "You are a discourse analysis expert. Analyze document structure, logical flow, and coherence in legal texts.", "model_name": "gpt-4o", "max_loops": 1, "temperature": 0.3 }, { "agent_name": "Risk Language Detector", "description": "Identifies risk-related language patterns", "system_prompt": "You are a legal risk language expert. Identify risk indicators, liability language, and potential legal concerns in contract language.", "model_name": "gpt-4o", "max_loops": 1, "temperature": 0.2 } ], "max_loops": 1 }'

=== "Python (requests)" ```python import requests import json

API_BASE_URL = "https://api.swarms.world"
API_KEY = "your_api_key_here"

headers = {
    "x-api-key": API_KEY,
    "Content-Type": "application/json"
}

swarm_config = {
    "name": "Legal Document Analysis MALT",
    "description": "Advanced linguistic analysis of legal documents using MALT framework",
    "swarm_type": "MALT",
    "task": "Perform comprehensive linguistic analysis of a complex legal contract including sentiment analysis, risk identification, clause categorization, and language complexity assessment",
    "agents": [
        {
            "agent_name": "Syntactic Analyzer",
            "description": "Analyzes sentence structure and grammar",
            "system_prompt": "You are a syntactic analysis expert. Analyze sentence structure, grammatical patterns, and linguistic complexity in legal texts.",
            "model_name": "gpt-4o",
            "max_loops": 1,
            "temperature": 0.2
        },
        {
            "agent_name": "Semantic Analyzer",
            "description": "Analyzes meaning and semantic relationships",
            "system_prompt": "You are a semantic analysis expert. Extract meaning, identify semantic relationships, and analyze conceptual content in legal documents.",
            "model_name": "gpt-4o",
            "max_loops": 1,
            "temperature": 0.3
        },
        {
            "agent_name": "Pragmatic Analyzer",
            "description": "Analyzes context and implied meanings",
            "system_prompt": "You are a pragmatic analysis expert. Analyze contextual meaning, implied obligations, and pragmatic implications in legal language.",
            "model_name": "gpt-4o",
            "max_loops": 1,
            "temperature": 0.4
        },
        {
            "agent_name": "Discourse Analyzer",
            "description": "Analyzes document structure and flow",
            "system_prompt": "You are a discourse analysis expert. Analyze document structure, logical flow, and coherence in legal texts.",
            "model_name": "gpt-4o",
            "max_loops": 1,
            "temperature": 0.3
        },
        {
            "agent_name": "Risk Language Detector",
            "description": "Identifies risk-related language patterns",
            "system_prompt": "You are a legal risk language expert. Identify risk indicators, liability language, and potential legal concerns in contract language.",
            "model_name": "gpt-4o",
            "max_loops": 1,
            "temperature": 0.2
        }
    ],
    "max_loops": 1
}

response = requests.post(
    f"{API_BASE_URL}/v1/swarm/completions",
    headers=headers,
    json=swarm_config
)

if response.status_code == 200:
    result = response.json()
    print("MALT framework completed successfully!")
    print(f"Linguistic analysis: {result['output']['linguistic_analysis']}")
    print(f"Cost: ${result['metadata']['billing_info']['total_cost']}")
    print(f"Execution time: {result['metadata']['execution_time_seconds']} seconds")
else:
    print(f"Error: {response.status_code} - {response.text}")
```

Example Response:

{
  "status": "success",
  "swarm_name": "legal-document-analysis-malt",
  "swarm_type": "MALT",
  "task": "Perform comprehensive linguistic analysis of a complex legal contract including sentiment analysis, risk identification, clause categorization, and language complexity assessment",
  "output": {
    "linguistic_analysis": {
      "syntactic_analysis": {
        "complexity_score": 8.2,
        "sentence_structure": "Predominantly complex and compound-complex sentences",
        "grammatical_patterns": "Heavy use of passive voice, subordinate clauses, and technical terminology",
        "readability": "Graduate level (16+ years of education required)"
      },
      "semantic_analysis": {
        "key_concepts": ["liability", "indemnification", "force majeure", "intellectual property"],
        "semantic_relationships": "Strong hierarchical concept relationships with clear definitional structures",
        "conceptual_density": "High - 3.2 legal concepts per sentence average",
        "ambiguity_indicators": ["potentially", "reasonable efforts", "material adverse effect"]
      },
      "pragmatic_analysis": {
        "implied_obligations": [
          "Good faith performance expected",
          "Timely notice requirements implied",
          "Mutual cooperation assumed"
        ],
        "power_dynamics": "Balanced with slight advantage to service provider",
        "speech_acts": "Predominantly commissives (commitments) and directives (obligations)"
      },
      "discourse_analysis": {
        "document_structure": "Well-organized with clear section hierarchy",
        "logical_flow": "Sequential with appropriate cross-references",
        "coherence_score": 8.5,
        "transition_patterns": "Formal legal transitions with clause numbering"
      },
      "risk_language": {
        "high_risk_terms": ["unlimited liability", "personal guarantee", "joint and several"],
        "risk_mitigation_language": ["subject to", "limited to", "except as provided"],
        "liability_indicators": 23,
        "risk_level": "Medium-High"
      }
    },
    "comprehensive_summary": {
      "language_complexity": "High complexity legal document requiring specialized knowledge",
      "risk_assessment": "Medium-high risk with standard legal protections",
      "readability_concerns": "Requires legal expertise for full comprehension",
      "recommendations": [
        "Consider plain language summary for key terms",
        "Review unlimited liability clauses",
        "Clarify ambiguous terms identified"
      ]
    }
  },
  "metadata": {
    "malt_framework": {
      "linguistic_layers_analyzed": 5,
      "language_processing_depth": "Advanced multi-layer analysis",
      "specialized_nlp_operations": [
        "Syntactic parsing",
        "Semantic role labeling", 
        "Pragmatic inference",
        "Discourse segmentation",
        "Risk pattern recognition"
      ]
    },
    "execution_time_seconds": 35.7,
    "billing_info": {
      "total_cost": 0.089
    }
  }
}

Best Practices

Use MALT for sophisticated language processing tasks
Design agents with complementary linguistic analysis capabilities
Ideal for tasks requiring deep language understanding
Consider multiple levels of linguistic analysis (syntax, semantics, pragmatics)

SequentialWorkflow - For ordered language processing
MixtureOfAgents - For diverse linguistic expertise
HierarchicalSwarm - For structured language analysis

10 KiB Raw Blame History