Apply node.input to the Step of Execution and simulateTask

[chrisli30]

This pull request introduces enhancements to the task engine's handling of input data, improves debugging capabilities, and refactors variable management for workflow nodes. The changes primarily focus on supporting dual-access mappings for input and output data, refining logging mechanisms, and ensuring backward compatibility with existing workflows.

Enhancements to input data handling:

• Added extraction and processing of trigger input data in SimulateTask, RunTask, and RunTriggerRPC, enabling dual-access mappings for better input data management. (core/taskengine/engine.go [1] [2] [3]; core/taskengine/executor.go [4] [5] [6]; core/taskengine/run_node_immediately.go [7] [8]

Improvements to debugging and logging:

• Enhanced collectInputKeysForLog to exclude current node variables and added detailed debug logs for better traceability during execution. (core/taskengine/vm.go core/taskengine/vm.goL1518-R1648)
• Included input data in execution logs for various node types (e.g., REST API, GraphQL, contract reads/writes, custom code, etc.) to facilitate debugging. (core/taskengine/vm.go [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]

Refactoring variable management:

• Introduced SetInputVarForStep to manage input variables for workflow nodes, ensuring dual-access mappings and backward compatibility. (core/taskengine/vm.go [1] [2]
• Refactored variable initialization and updates to support both data and input fields for nodes and triggers. (core/taskengine/vm.go [1] [2]; core/taskengine/engine.go [3] [4]; core/taskengine/executor.go [5] [6]

Code cleanup and dependency updates:

• Updated imports to include structpb for handling protobuf values and reorganized imports for better readability. (core/taskengine/engine.go [1]; core/taskengine/executor.go [2]; core/taskengine/vm.go [3]

These changes collectively improve the task engine's flexibility, debugging capabilities, and compatibility with complex workflows.

[Copilot]

Pull Request Overview

This PR extends the system to capture and propagate user-provided input data (node.input and trigger.input) throughout execution and simulation. It adds new input fields to protobuf messages, updates GRPC method signatures, and enhances the VM and per-node processors to set and log those inputs.

• Added input fields to all trigger and node protobuf messages and updated RPC definitions.
• Updated VM (vm.go) and each *Processor.Execute to extract, store, and include input data in Execution_Step.
• Modified collectInputKeysForLog to skip the current node and emit both .data and .input entries.

Reviewed Changes

Copilot reviewed 16 out of 16 changed files in this pull request and generated 1 comment.

File	Description
protobuf/avs.proto	Added input fields to trigger and node messages and updated RPC request.
core/taskengine/vm_runner_rest.go	Extract and include node.Input in Execution_Step.
core/taskengine/vm.go	New ExtractNodeInputData, SetInputVarForStep, enhanced input propagation and collectInputKeysForLog.

Comments suppressed due to low confidence (1)

core/taskengine/vm.go:819

• There are no tests validating that Execution_Step.Input is correctly set or that collectInputKeysForLog correctly excludes the current node. Adding unit tests for these new behaviors will help catch regressions.

// Extract and set input data for this node (making it available as node_name.input)