Skip to content

workflow-guide.md

Latest commit

 

History

History
56 lines (42 loc) · 1.95 KB

workflow-guide.md

File metadata and controls

56 lines (42 loc) · 1.95 KB
id doc-039
title Workflow Guide
type other
created_date 2026-04-01 15:29

Workflow Guide

This document describes the core workflows for maintaining the site data, building the archive, and performing deployments.

Site Architecture

The site is built with Jekyll but is primarily data-driven.

  • _data/: Canonical YAML files for interviews, video assets, resumes, and taxonomy.
  • bin/: Generator scripts that transform YAML data into Jekyll-readable HTML files.
  • _site/: The rendered static site (historically committed to Git, but planned to be ignored in the modernization phase).

Core Workflows

1. Updating Resume Content

Resume content is stored in _data/resume/.

  • timeline.yml: Controls the order of positions.
  • positions/*.yml: Contains detailed role descriptions.
  • Changes to these files will automatically update the HTML, plain text, and Markdown resume formats.

2. Adding New Interviews

  1. Add the interview metadata to _data/interviews.yml.
  2. Add the video asset metadata to _data/video_assets.yml.
  3. If a transcript is available, add it to _data/transcripts/.
  4. Run ./bin/pipeline build to generate the new pages.

3. Build & Deployment

The canonical entry point for all operations is ./bin/pipeline.

# Full local CI build and validation
./bin/pipeline ci

# Run smoke tests
./bin/pipeline smoke

4. Data Validation

Data integrity is enforced by multiple scripts in the CI pipeline.

  • validate_data_integrity.rb: Checks for required fields and cross-references.
  • validate_data_uniqueness.rb: Ensures IDs and slugs are unique.
  • audit_transcripts.rb: Checks for missing or malformed transcripts.

Modernization Strategy

The project is currently transitioning to a more automated and robust architecture:

  • Moving build/deploy to GitHub Actions.
  • Using Jekyll memory-based generators instead of disk-based HTML generation.
  • Consolidating fragmented scripts into a unified CLI.