ConstraintRelaxationTask

⚙️ TaskConfig.json


{
  "task_type": "ConstraintRelaxation",
  "problem": "Design a high-performance web server for 1M req/s",
  "constraints": {
    "Latency < 10ms": 1.0,
    "Cost < $100/mo": 0.4,
    "Zero-Trust Security": 0.9
  },
  "relaxation_strategy": "progressive",
  "find_creative_satisfactions": true
}

→

👁️ Session UI (TabbedDisplay)

Overview Constraint Analysis Iteration 1 Final Synthesis

Progressive Reintroduction

✅ Iteration 1 complete: Zero-Trust Security (1.2s)

⏳ Adapting solution for: Latency < 10ms...

Creative Insight: Using eBPF for security checks to maintain latency targets.

Live Results Showcase

Explore actual artifacts generated by this task in our test workspace.

Execution Configuration

Field	Type	Description
`problem` *	String	The problem description to solve.
`constraints` *	Map<String, Double>	Map of constraints to priority weights (0.0-1.0, where 1.0 is critical).
`relaxation_strategy`	String	`progressive` (gradual), `selective` (subset), or `hierarchical` (levels).
`reintroduction_order`	String	`by_priority`, `by_difficulty`, or `by_dependency`.
`find_creative_satisfactions`	Boolean	Whether to actively seek creative ways to satisfy constraints. Default: `true`.
`max_iterations`	Int	Maximum number of relaxation/reintroduction iterations (1-10). Default: `5`.
`input_files`	List<String>	Glob patterns for files to use as input context (e.g. `*/.kt`).
`related_files`	List<String>	Additional files for context.

Kotlin Boilerplate

Direct instantiation for use within the Cognotik environment:


val task = ConstraintRelaxationTask(
    orchestrationConfig = config,
    planTask = ConstraintRelaxationTaskExecutionConfigData(
        problem = "Optimize database schema for write-heavy workload",
        constraints = mapOf(
            "ACID Compliance" to 1.0,
            "Sub-millisecond writes" to 0.8,
            "Minimal storage overhead" to 0.3
        ),
        relaxation_strategy = "progressive",
        input_files = listOf("src/main/resources/schema.sql")
    )
)

Embedded Invocation

Using the UnifiedHarness for headless or automated agentic coding:


import com.simiacryptus.cognotik.plan.tools.reasoning.ConstraintRelaxationTask.Companion.ConstraintRelaxation
import com.simiacryptus.cognotik.plan.tools.reasoning.ConstraintRelaxationTask.ConstraintRelaxationTaskExecutionConfigData
harness.runTask(
    taskType = ConstraintRelaxation,
    executionConfig = ConstraintRelaxationTaskExecutionConfigData(
        problem = "Design a scalable microservices architecture",
        constraints = mapOf("High Availability" to 1.0, "Low Latency" to 0.9),
        relaxation_strategy = "progressive"
    ),
    workspace = File("./my-project"),
    autoFix = true
)


val task = ConstraintRelaxationTask(
    orchestrationConfig = config,
    planTask = ConstraintRelaxationTaskExecutionConfigData(
        problem = "Optimize database schema for write-heavy workload",
        constraints = mapOf(
            "ACID Compliance" to 1.0,
            "Sub-millisecond writes" to 0.8,
            "Minimal storage overhead" to 0.3
        ),
        relaxation_strategy = "progressive",
        input_files = listOf("src/main/resources/schema.sql")
    )
)

Prompt Segment

The following logic is injected into the LLM context:


ConstraintRelaxation - Solve over-constrained problems through progressive constraint relaxation
  ** Specify the problem to solve
  ** Define constraints with priority weights (0.0-1.0, where 1.0 is critical)
  ** Choose relaxation strategy:
     - 'progressive': Gradually relax constraints from lowest to highest priority
     - 'selective': Intelligently select which constraints to relax
     - 'hierarchical': Relax constraints in priority-based levels
  ** Choose reintroduction order:
     - 'by_priority': Reintroduce highest priority constraints first
     - 'by_difficulty': Reintroduce easiest constraints first
     - 'by_dependency': Reintroduce based on constraint dependencies
  ** Enable creative satisfaction finding to discover novel solutions
  ** Produces a solution that progressively satisfies constraints
  ** Shows evolution of solution as constraints are reintroduced
  ** Optionally, list input files (supports glob patterns) to be examined for context

Progressive Reintroduction

Live Results Showcase

Execution Configuration

Task Lifecycle

1. Constraint Analysis

2. Initial Relaxed Solution

3. Iterative Reintroduction

4. Final Synthesis

Kotlin Boilerplate

Embedded Invocation

Prompt Segment