Definition

Isomorphism of Intent is the property where a behavioral specification (expressed in natural language, Gherkin, or visual diagrams) maintains a one-to-one correspondence with its deterministic execution in a runtime system.

In formal terms: Given a specification S and an execution trace E, an agentic workflow exhibits isomorphism of intent when:

∀ s ∈ S, ∃! e ∈ E : φ(s) = ψ(e)

Where:

• φ: S → B maps specifications to behaviors
• ψ: E → B maps executions to behaviors
• B is the space of observable behaviors
• The mapping is bijective (one-to-one and onto)

The Problem It Solves

Traditional LLM-based automation suffers from the Intentionality Gap: the fidelity loss between what a human specifies and what the system executes.

• Chatbots: Probabilistic responses, no guarantees
• Prompt Engineering: Unstable, context-dependent
• Critical Systems (QA, Drones, Finance): Cannot tolerate fidelity loss

Why This Matters

When you specify “if payment fails, retry with exponential backoff”, you don’t want probabilistic interpretation. You want deterministic execution.

Isomorphism of Intent ensures that specification and execution are mathematically equivalent—no interpretation gap.

Implementation

This is achieved through:

1. Semantic Extraction: Parse intent into a canonical form (DAG)
2. Orthogonal Orchestration: Separate behavior (Gherkin) from interface (Figma)
3. State Invariance Verification: Prove execution maintains specification constraints

Related Concepts

• [[Orthogonal Orchestration]]
• [[Intentionality Gap]]
• [[Semantic Execution]]