Assignment Action Roadmap#

This document outlines the strategic roadmap for the Assignment action type, moving from basic state mutation to a robust, type-safe, and highly efficient data manipulation engine.

1. Summary of Issues & Gaps#

The current Assignment implementation (v1) provides fundamental batch mutation capabilities but faces several limitations when applied to complex enterprise orchestration:

• Shallow Mutation: No native support for deep document paths (e.g., doc.items.0.qty). Currently requires a separate Loop node.
• Limited Scope: Mutations are restricted to the current document (doc) and local variables (vars). No cross-document mutation support.
• Implicit Typing: Operators like increment perform runtime type coercion (float conversion) but lack proactive schema validation.
• Trigger Recursion Risks: Basic loop detection exists, but there is no specialized “mutation-aware” cycle prevention (e.g., field-level change suppression).
• Concurrency & State: No state snapshotting or “Revert on Failure” mechanics at the individual action row level.

2. Short-Term Fixes (v1 Improvements)#

Goal: Improve stability, safety, and basic usability.

Engine Enhancements#

• Initialization Guard (High): Automatically initialize vars to a default value (0, empty string, etc.) if they don’t exist during an increment or append operation to avoid manual initialization rows.
• Field-Level Suppress (Medium): Add a flag to skip change-tracking hooks for specific assignments to prevent triggering recursive rules.
• Strict Boolean Toggle (Low): Update Toggle operator to strictly enforce boolean input, raising a clear validation error for non-boolean types instead of best-effort coercion.

UI / Builder Enhancements#

• Inline Type Hints (High): Display the target field’s data type (Int, Check, Link, etc.) in the Assignment row to help users choose the correct operator.
• Resolver Auto-Complete (Medium): Provide suggestions for available doc and vars paths within the Value Resolver text area.

3. Mid-Term Enhancements (Engine + UX)#

Goal: Expand functional reach and developer observability.

Engine Enhancements#

• Deep Path Support (High): Implement dot-notation support for nested dictionaries and child tables (e.g., doc.items.last.rate or doc.table_name.row_id.field).
• Conditional Merge Strategies (Medium): Add “Deep Merge” and “Overwrite if Null” operators for JSON and Dictionary targets.
• Safe Math Operations (Low): New operators for Multiply By, Divide By, and Round To, reducing reliance on complex Formulas for basic math.

UI / Builder Enhancements#

• Mutation Preview (High): A “Simulation Mode” in the builder that shows a before/after diff of the document based on the current Assignment configuration.
• Execution Trace Visualization (Medium): Enhanced visual logs showing the value resolution for every single row in a batch assignment.

4. Long-Term Vision (Rule Engine Evolution)#

Goal: Transform Assignment into a declarative data transform layer.

Functional Evolution#

• Cross-Document Mutations (High): Support for link.TARGET_DOC.field targets, allowing a rule on an Invoice to directly update its parent Sales Order.
• State Snapshots & Atomic Rollbacks (Medium): The ability to “checkpoint” document state before a batch assignment and perform a partial rollback of just that node on failure.
• Reactive Trigger Suppression (Low): Intelligence to detect and block “Circular Assignment Paths” at compile-time (e.g., doc.A updates doc.B, which has a rule that updates doc.A).

5. Summary Table#