coding-workflow

Batch document coding orchestrates systematic processing of document groups through Stage 2 qualitative analysis, automatically tracking coding progress and generating comprehensive audit trails. Initialize your project, complete Stage 1, then run process-batch.js with your document list and assigned agent to begin.

Stage 1 establishes initial data structure and project setup. Stage 2 automates batch processing, document tracking across status states (pending, in_progress, coded, reviewed, finalized), and creates detailed audit documentation logging all coding activities and AI interactions for transparency.

Yes, document tracking monitors each document's status throughout the workflow and persists across sessions. The audit trail logs all coding activities, human decisions, and AI interactions, enabling you to resume batch sessions and generate comprehensive methods documentation.

Yes, Stage 1 completion is a prerequisite. You must establish initial data structure, initialize your project, and pass readiness checks via read-config.js before Stage 2 batch processing can orchestrate document coding and audit documentation.

Pass your document IDs and target agent (e.g., @dialogical-coder) to process-batch.js with the --documents and --agent flags. Specify the coding phase (e.g., phase2_synthesis) to route your batch through the designated agent for systematic analysis.

The audit trail documents all coding activities, human decisions, and AI agent interactions with full transparency. It logs document status transitions, coding actions, and decisions throughout the workflow, producing confirmable records for your research methods section.