CtrlK

Community Documentation Log in Get started

tessl/maven-com-embabel-agent--embabel-agent-domain

Core domain type definitions for the Embabel Agent Framework, providing foundational data classes and interfaces for agent-based AI workflows including content assets, research entities, and person types with Jackson serialization and PromptContributor capabilities.

Overview

Eval results

Files

Research Types

Name: tessl/maven-com-embabel-agent--embabel-agent-domain
Author: tessl

Domain models for research workflows in agent systems.

ResearchTopic

Represents a research topic with specific questions to investigate.

@JsonClassDescription("topic to research")
open class ResearchTopic(
    @get:JsonPropertyDescription("topic to research")
    val topic: String,
    @get:JsonPropertyDescription("specific questions")
    val questions: List<String>
)

Properties:

topic: String - Research topic (e.g., "Quantum Computing", "AI Ethics")
questions: List<String> - Specific questions to investigate

Methods:

toString(): String - Format: "ResearchTopic(topic='...', questions=[...])"

Note: Open class - can be extended with additional properties.

ResearchTopics

Container for multiple research topics.

open class ResearchTopics(
    val topics: List<ResearchTopic>
)

Properties:

topics: List<ResearchTopic> - List of research topics

Note: Open class - can be extended.

ResearchReport

Comprehensive research report with content and linked internet resources.

@JsonClassDescription("Research report, containing a topic, content and links")
open class ResearchReport(
    @get:JsonPropertyDescription("topic of the report")
    val topic: String,
    @get:JsonPropertyDescription("content of the report")
    override val content: String,
    @get:JsonPropertyDescription("links to internet resources")
    override val links: List<InternetResource>
) : InternetResources, HasInfoString, ContentAsset

Properties:

topic: String - Report topic (e.g., "Quantum Computing", "AI Ethics")
content: String - Report text (from HasContent via ContentAsset)
links: List<InternetResource> - Referenced internet resources (from InternetResources)
timestamp: Instant - Creation timestamp (from Timestamped via ContentAsset), auto-set to Instant.now()

Interfaces:

InternetResources - Provides links property and contribution() method
HasInfoString - Provides infoString() method
ContentAsset - Provides content, timestamp, and LLM integration

Methods:

infoString(verbose: Boolean? = null, indent: Int = 0): String - Formatted info string with content and links
toString(): String - Returns infoString(verbose = false, indent = 0)
contribution(): String - Formatted report for LLM prompts

Note: Open class - can be extended.

Basic Usage

Creating ResearchTopic

import com.embabel.agent.domain.library.ResearchTopic

val topic = ResearchTopic(
    topic = "AI Ethics in Modern Systems",
    questions = listOf(
        "What are key ethical concerns in AI deployment?",
        "How can we mitigate algorithmic bias?",
        "What frameworks exist for ethical AI development?"
    )
)

println(topic.topic)      // "AI Ethics in Modern Systems"
println(topic.questions)  // [What are key..., How can we..., What frameworks...]

Creating ResearchTopics

import com.embabel.agent.domain.library.ResearchTopics
import com.embabel.agent.domain.library.ResearchTopic

val topics = ResearchTopics(
    topics = listOf(
        ResearchTopic(
            topic = "Machine Learning",
            questions = listOf("What are latest developments?")
        ),
        ResearchTopic(
            topic = "Distributed Systems",
            questions = listOf("How to ensure consistency?")
        )
    )
)

// Access topics
topics.topics.forEach { topic ->
    println("Topic: ${topic.topic}")
    topic.questions.forEach { question ->
        println("  - $question")
    }
}

Creating ResearchReport

import com.embabel.agent.domain.library.InternetResource
import com.embabel.agent.domain.library.ResearchReport
import java.time.Instant

val report = ResearchReport(
    topic = "AI Ethics in Healthcare",
    content = """
        # AI Ethics in Healthcare

        ## Executive Summary
        AI systems in healthcare must balance innovation with patient safety...

        ## Key Findings
        1. Bias in diagnostic algorithms affects marginalized populations
        2. Data privacy concerns are paramount
        3. Transparency in AI decision-making is essential

        ## Recommendations
        - Implement rigorous testing for algorithmic bias
        - Establish clear data governance policies
        - Require explainable AI for critical decisions
    """.trimIndent(),
    links = listOf(
        InternetResource(
            url = "https://www.nature.com/ai-healthcare-ethics",
            summary = "Comprehensive study on AI ethics in healthcare"
        ),
        InternetResource(
            url = "https://www.who.int/ai-ethics-guidelines",
            summary = "WHO guidelines for ethical AI"
        ),
        InternetResource(
            url = "https://arxiv.org/abs/bias-diagnostic-ai",
            summary = "Research on algorithmic bias in diagnostics"
        )
    )
)

// Access properties
println(report.topic)      // "AI Ethics in Healthcare"
println(report.content)    // Markdown content
println(report.timestamp)  // Instant when created
println(report.links)      // List of InternetResource

ResearchReport Methods

val report = ResearchReport(
    topic = "Quantum Computing Applications",
    content = "Quantum computing has emerged...",
    links = listOf(
        InternetResource(
            url = "https://quantum.ibm.com/applications",
            summary = "IBM quantum applications"
        ),
        InternetResource(
            url = "https://www.nature.com/quantum-applications",
            summary = "Nature article on quantum apps"
        )
    )
)

// Info string (for logging/debugging)
val info = report.infoString(verbose = false, indent = 0)
// Returns:
// |Report:
// |Quantum computing has emerged...
// |Links: [urls]

// toString uses infoString
println(report.toString())

// LLM contribution
val contribution = report.contribution()
// Returns:
// Research Report:
// Topic: Quantum Computing Applications
// Content: Quantum computing has emerged...
// Links: [url - summary, ...]
// Date: 2026-02-06

LLM Integration

val report = ResearchReport(
    topic = "AI Agent Frameworks",
    content = """
        # Framework Landscape

        ## Overview
        The ecosystem has evolved significantly...

        ## Leading Frameworks
        1. Embabel - JVM-based, sophisticated planning
        2. LangChain - Python/JS, wide ecosystem
        3. AutoGen - Microsoft, multi-agent focus
    """.trimIndent(),
    links = listOf(
        InternetResource(
            url = "https://github.com/embabel/embabel-agent",
            summary = "Embabel repository"
        ),
        InternetResource(
            url = "https://www.langchain.com/",
            summary = "LangChain docs"
        )
    )
)

val promptContext = """
    Research Context:
    ${report.contribution()}

    Based on this, recommend...
""".trimIndent()

Extending Types

Extended ResearchTopic

import java.time.Instant

class PrioritizedResearchTopic(
    topic: String,
    questions: List<String>,
    val priority: Int,
    val deadline: Instant?
) : ResearchTopic(topic, questions) {

    override fun toString(): String =
        "PrioritizedResearchTopic(topic='$topic', questions=$questions, " +
        "priority=$priority, deadline=$deadline)"
}

val urgentTopic = PrioritizedResearchTopic(
    topic = "Security Vulnerabilities",
    questions = listOf("Current threats?", "Patching strategy?"),
    priority = 1,
    deadline = Instant.now().plusSeconds(86400)
)

Extended ResearchReport

class AnnotatedResearchReport(
    topic: String,
    content: String,
    links: List<InternetResource>,
    val annotations: List<String>,
    val reviewers: List<String>
) : ResearchReport(topic, content, links) {

    override fun contribution(): String =
        super.contribution() +
        "\nAnnotations: ${annotations.joinToString("; ")}"
}

val annotated = AnnotatedResearchReport(
    topic = "Security Best Practices",
    content = "The following practices...",
    links = listOf(),
    annotations = listOf("Needs TLS 1.3 update", "Add zero-trust section"),
    reviewers = listOf("Alice", "Bob")
)

JSON Serialization

import com.fasterxml.jackson.module.kotlin.jacksonObjectMapper
import com.fasterxml.jackson.datatype.jsr310.JavaTimeModule

val mapper = jacksonObjectMapper().registerModule(JavaTimeModule())

// ResearchTopic
val topic = ResearchTopic(
    topic = "Quantum Computing",
    questions = listOf("What is quantum supremacy?", "Practical applications?")
)

val topicJson = mapper.writeValueAsString(topic)
val topicDeserialized = mapper.readValue(topicJson, ResearchTopic::class.java)

// ResearchReport
val report = ResearchReport(
    topic = "Cloud Security",
    content = "Cloud security encompasses...",
    links = listOf(
        InternetResource(
            url = "https://cloud.google.com/security",
            summary = "GCP security best practices"
        )
    )
)

val reportJson = mapper.writeValueAsString(report)
val reportDeserialized = mapper.readValue(reportJson, ResearchReport::class.java)

ResearchTopic JSON Format:

{
  "topic": "Quantum Computing",
  "questions": [
    "What is quantum supremacy?",
    "What are practical applications?"
  ]
}

ResearchReport JSON Format:

{
  "topic": "Cloud Security",
  "content": "Cloud security encompasses...",
  "links": [
    {
      "url": "https://cloud.google.com/security",
      "summary": "GCP security best practices"
    }
  ],
  "timestamp": "2026-02-06T12:00:00Z"
}

Complete Workflow

import com.embabel.agent.domain.library.*

// 1. Define research topics
val topics = ResearchTopics(
    topics = listOf(
        ResearchTopic(
            topic = "AI Agent Frameworks",
            questions = listOf(
                "What are leading frameworks?",
                "How do they compare?",
                "What are adoption trends?"
            )
        )
    )
)

// 2. Conduct research (in agent action)
// ...

// 3. Create research report
val report = ResearchReport(
    topic = "AI Agent Frameworks",
    content = """
        # Framework Landscape

        ## Overview
        The ecosystem has evolved...

        ## Leading Frameworks
        1. Embabel - JVM, planning
        2. LangChain - Python/JS
        3. AutoGen - Multi-agent

        ## Comparison
        [Detailed comparison...]
    """.trimIndent(),
    links = listOf(
        InternetResource(
            url = "https://github.com/embabel/embabel-agent",
            summary = "Embabel framework"
        ),
        InternetResource(
            url = "https://www.langchain.com/",
            summary = "LangChain docs"
        ),
        InternetResource(
            url = "https://microsoft.github.io/autogen/",
            summary = "Microsoft AutoGen"
        )
    )
)

// 4. Use in LLM prompts
val prompt = """
    Research:
    ${report.contribution()}

    Recommend...
""".trimIndent()

Integration with Other Types

With Summary

fun summarizeResearch(report: ResearchReport): Summary {
    return Summary(
        "Research on '${report.topic}': ${report.content.take(200)}... " +
        "References ${report.links.size} sources."
    )
}

With Blog

fun createBlogFromResearch(report: ResearchReport, author: String): Blog {
    val linksSection = report.links.joinToString("\n") { link ->
        "- [${link.summary}](${link.url})"
    }

    val content = """
        ${report.content}

        ## References
        $linksSection
    """.trimIndent()

    return Blog(
        title = report.topic,
        author = author,
        content = content,
        keywords = setOf("research", report.topic.lowercase())
    )
}

With News

fun enrichResearchWithNews(
    report: ResearchReport,
    news: RelevantNewsStories
): String {
    return """
        ${report.contribution()}

        ## Recent News
        ${news.contribution()}
    """.trimIndent()
}

With Person

data class Researcher(
    override val name: String,
    val affiliation: String
) : Person

val researchers = listOf(
    Researcher("Dr. Alice", "MIT"),
    Researcher("Dr. Bob", "Stanford")
)

val report = ResearchReport(
    topic = "Collaborative AI",
    content = """
        Research by: ${researchers.joinToString { "${it.name} (${it.affiliation})" }}

        Findings...
    """.trimIndent(),
    links = listOf()
)

tessl i tessl/maven-com-embabel-agent--embabel-agent-domain@0.3.0

docs

integration-patterns.md

json-serialization.md

tessl/maven-com-embabel-agent--embabel-agent-domain

research-types.md.css-3qkkll{font-size:var(--chakra-font-sizes-sm);font-weight:var(--chakra-font-weights-normal);color:var(--chakra-colors-gray-300);}docs/

Research Types

ResearchTopic

ResearchTopics

ResearchReport

Basic Usage

Creating ResearchTopic

Creating ResearchTopics

Creating ResearchReport

ResearchReport Methods

LLM Integration

Extending Types

Extended ResearchTopic

Extended ResearchReport

JSON Serialization

Complete Workflow

Integration with Other Types

With Summary

With Blog

With News

With Person

research-types.mddocs/