Add merge-doc pipeline node
Objective
Convert annotated transcript into final human-editable document structure.
Why this matters
This is the final transformation stage before human editing and export.
It becomes the canonical “source of truth” transcript.
Parallelization metadata
Track
pipeline
Depends on
Safe to run in parallel with
- pipeline validation
- remotion cleanup
Merge risk
Files in scope
Primary:
scripts/pipeline/nodes/merge-doc.ts
Required implementation
1. Implement merge node
Wrap existing merge logic.
2. Inputs
Must accept:
- speaker-assigned transcript
- doc directives
3. Output
Produces:
Constraints
- no UI coupling
- no editor logic
Acceptance criteria
Functional
Integration checks
Verification commands
Run pipeline through merge-doc.
Expected result:
- final transcript artifact generated
Explicitly out of scope
- editor UI
- version history
Suggested branch
refactor/s7-merge-doc
Suggested commit slug
phase-2-step-7-merge-doc
Add
merge-docpipeline nodeObjective
Convert annotated transcript into final human-editable document structure.
Why this matters
This is the final transformation stage before human editing and export.
It becomes the canonical “source of truth” transcript.
Parallelization metadata
Track
pipelineDepends on
alignpipeline node #41assign-speakerspipeline node #46Safe to run in parallel with
Merge risk
Files in scope
Primary:
scripts/pipeline/nodes/merge-doc.tsRequired implementation
1. Implement merge node
Wrap existing merge logic.
2. Inputs
Must accept:
3. Output
Produces:
Constraints
Acceptance criteria
Functional
Integration checks
Verification commands
Run pipeline through merge-doc.
Expected result:
Explicitly out of scope
Suggested branch
refactor/s7-merge-docSuggested commit slug
phase-2-step-7-merge-doc