Cognotik | DialecticalReasoningTask

ExecutionConfig.json JSON

{
  "thesis": "Strict microservices for scale",
  "antithesis": "Monolithic simplicity",
  "context": "E-commerce Platform Migration",
  "synthesis_levels": 3,
  "preserve_strengths": true
}

→

Session UI (TabbedDisplay) Rendered

Overview

Thesis

Antithesis

Synthesis L3

Synthesis Level 3: Macro-Service Mesh

The final synthesis transcends the scale/simplicity trade-off by proposing a "Cell-Based Architecture" where logical monoliths are deployed as isolated scaling units...

Preserves: Local development speed
Preserves: Independent deployment
Resolves: Distributed transaction overhead

✔ Final Integration Complete (14.2s)

Test Workspace Browser

Explore generated dialectical reports and analysis artifacts in the task workspace.

Configuration Parameters

Field	Type	Description
`thesis`*	String	The primary position or claim to be analyzed. Max 5000 chars.
`antithesis`*	String	The opposing position or contradiction. Max 5000 chars.
`context`	String	Domain or background info to ground the analysis.
`synthesis_levels`	Int	Iterations of synthesis (1-5). Default: `3`.
`preserve_strengths`	Boolean	If true, ensures synthesis keeps valid points from both sides.
`input_files`	List<String>	Glob patterns for source files to include as primary context.
`related_files`	List<String>	Additional files for background context.

Task Lifecycle

Initialization: Validates that thesis and antithesis are distinct and within length limits.
Thesis/Antithesis Analysis: Independent agents analyze the core claims, strengths, and blind spots of each position.
Contradiction Mapping: Identifies the root causes of the tension and areas of hidden overlap.
Iterative Synthesis:
- Level 1: Finds a perspective that integrates both insights.
- Level 2+: Treats the previous synthesis as a new thesis to find even deeper abstractions.
Final Integration: Produces a comprehensive report with practical implications and actionable recommendations.

Embedded Execution

Invoke this task programmatically using the UnifiedHarness in serverless mode.

import com.simiacryptus.cognotik.plan.tools.social.DialecticalReasoningTask
import com.simiacryptus.cognotik.plan.tools.social.DialecticalReasoningTask.Companion.DialecticalReasoning

// 1. Define the Job Configuration
val config = DialecticalReasoningTask.DialecticalReasoningTaskExecutionConfigData(
    thesis = "Strict microservices for scale",
    antithesis = "Monolithic simplicity",
    context = "E-commerce Platform Migration",
    synthesis_levels = 3,
    preserve_strengths = true,
    input_files = listOf("src/main/kotlin/architecture/*.kt")
)

// 2. Run via Harness
harness.runTask(
    taskType = DialecticalReasoning,
    typeConfig = TaskTypeConfig(name = "ArchReview"),
    executionConfig = config,
    workspace = File("./project-root"),
    autoFix = true
)

CLI / CI Integration

Example GitHub Action step for automated reasoning:

- name: Run Dialectical Analysis
  run: |
    java -jar cognotik-cli.jar \
      --task DialecticalReasoning \
      --thesis "Implement custom auth" \
      --antithesis "Use Auth0 provider" \
      --workspace .
  env:
    OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}

Prompt Segment

The following logic is injected into the LLM orchestrator:

DialecticalReasoning - Resolve contradictions through thesis-antithesis-synthesis
  ** Specify thesis and antithesis statements representing opposing positions
  ** Provide context to ground the dialectical analysis
  ** Configure synthesis_levels (1-5) to iterate toward higher understanding
  ** Set preserve_strengths=true to maintain valuable aspects of both sides
 ** Related files can provide additional context
  ** Explores contradictions and tensions between positions
  ** Generates synthesis that transcends opposition
  ** Iterates to progressively higher levels of understanding
  ** Produces structured dialectical analysis with final synthesis