Employment Law Aid – Content Publishing Process

Purpose: Prevent sync errors and link issues by following a repeatable process from drafting in Obsidian to publishing on WordPress.

Quick refs

• WordPress (Local): /Users/cosmodrome/Local Sites/employment-law-aid
• Scripts: /Users/cosmodrome/Local Sites/employment-law-aid/scripts
• Obsidian Vault Project: /Users/cosmodrome/Library/Mobile Documents/iCloud~md~obsidian/Documents/Second Brain/8 - Projects/Employment Law Aid

0) Prerequisites

• WP-CLI installed and working for the local site.
• Frontmatter on all content follows project schema at minimum:
  • slug, title, description, status (Draft or Published), optional: type, funnel_stage, related_pages, target_keywords.
• Slug conventions
  • National/topic hubs: eeoc, eeoc/deadlines, eeoc/right-to-sue, etc.
  • Hierarchies: eeoc/offices/california (never encode the full path into post_name; use parent-child).

0.5) Keyword research workflow (inputs and steps)

Primary inputs

• CSV export: /Users/cosmodrome/Downloads/organic-keywords-employmentlawhelp_org.csv
• Internal note: /Users/cosmodrome/Library/Mobile Documents/iCloud~md~obsidian/Documents/Second Brain/8 - Projects/Employment Law Aid/Competitor Keyword Opportunities - Priority List.md

Steps

1. Review the Priority List note (Executive Summary + Priority Queue) to set near-term targets.
2. Scan the CSV for high-volume, low-difficulty clusters not yet covered in the vault (difficulty ≤10 as quick wins).
3. Map keywords → proposed slugs using existing URL conventions and hierarchy.
4. Check for existing pages by slug to avoid duplicates.
5. Add target_keywords to frontmatter when drafting new pages.

Optional helpers

• Quick check for proposed slug existence (replace list with candidates):

python3 - <<'PY'
import os, re, json
ROOT = "/Users/cosmodrome/Library/Mobile Documents/iCloud~md~obsidian/Documents/Second Brain/8 - Projects/Employment Law Aid"
FM = re.compile(r"^---\n(.*?)\n---", re.S); FLD = re.compile(r"^([A-Za-z0-9_\-]+):\s*(.*)$")
slugs = {}
for root,_,files in os.walk(ROOT):
  for fn in files:
    if not fn.endswith('.md'): continue
    p=os.path.join(root,fn); t=open(p,'r',encoding='utf-8').read(); m=FM.match(t)
    if not m: continue
    fm={}
    for line in m.group(1).splitlines():
      m2=FLD.match(line.strip() or '')
      if m2:
        k,v=m2.group(1),m2.group(2).strip().strip('"').strip("'")
        if k=='slug': slugs[v.strip('/ ')] = p
print(json.dumps({s: slugs.get(s) for s in [
  'north-carolina/at-will-employment-getting-fired',
  'colorado/at-will-employment-getting-fired',
  'oregon/at-will-employment-getting-fired',
  'new-york/wrongful-termination/warn-act',
  'florida/wages-and-hours/meal-break-laws',
  'michigan/wages-and-hours/meal-break-laws',
  'pennsylvania/wages-and-hours/meal-break-laws',
  'discrimination/workplace-discrimination-examples',
  'resources/sample-employment-complaint-letters'
]}, indent=2))
PY

1) Content prep in Obsidian

• Linking policy in the vault:
  • Use standard markdown links (not wiki-links) while drafting.
  • Prefer relative .md links between notes during editing for readability.
  • Before syncing to WordPress, convert any .md links to site URLs based on the target note’s slug (see Link conversion step).
• Frontmatter
  • Ensure every note has a slug and correct status.

2) Build/refresh slug index

Regenerate the slug map used by link tools and QA.

python3 tools/build_slug_index.py

3) Create missing WordPress pages (hierarchical)

Creates pages according to slug, building parents first and adding taxonomy.

python3 "/Users/cosmodrome/Local Sites/employment-law-aid/scripts/create_missing_pages.py"

Notes

• Existence is checked by (post_parent, post_name) pairs.
• Content-level taxonomy is auto-assigned by depth:
  • Depth 0–1: content-level = topic
  • Depth ≥2: content-level = l3-spoke

4) Link conversion (relative .md → site URLs)

Convert remaining .md links to their final site URLs using the slug index.

Run the link rewriter:

python3 "/Users/cosmodrome/Local Sites/employment-law-aid/scripts/rewrite_internal_links.py" \
  --md-dir "/Users/cosmodrome/Library/Mobile Documents/iCloud~md~obsidian/Documents/Second Brain/8 - Projects/Employment Law Aid" \
  --only-prefix eeoc

Options

• Add --include-html to convert HTML anchors as well.
• Omit --only-prefix to process the entire vault.
• Use --dry-run to preview changes.

After conversion, all internal links in markdown should look like /eeoc/deadlines/ (absolute site paths), not ./EEOC - Deadlines (...).md.

5) Sync content to WordPress (safe)

Uploads post_content + ACF H1 + Rank Math meta for notes with status: Draft.

python3 "/Users/cosmodrome/Local Sites/employment-law-aid/scripts/sync_published_safe.py"

Notes

• Script finds the WP page by hierarchical slug matching.
• On successful upload, it updates the note’s status to Published.

6) Validate internal links against the site

Crawl internal links and ensure no 404s.

python3 "/Users/cosmodrome/Local Sites/employment-law-aid/scripts/validate_links.py" \
  --md-dir "/Users/cosmodrome/Library/Mobile Documents/iCloud~md~obsidian/Documents/Second Brain/8 - Projects/Employment Law Aid" \
  --base-url "http://employment-law-aid.local"

If broken links are reported

• Fix the source markdown (usually lingering .md links or bad filename mappings).
• Re-run the sync script to push corrected content.

7) Duplicate/flattened page cleanup (as needed)

If a script created flat pages like eeoc-offices-new-york at the top level, remove them.

List candidates (example for eeoc- prefix):

wp --path="/Users/cosmodrome/Local Sites/employment-law-aid/app/public" --no-color \
  post list --post_type=page --fields=ID,post_name,post_parent --format=json \
  | jq '[.[] | select((.post_parent==0) and (.post_name|startswith("eeoc-")) and (.post_name!="eeoc")) ]'

Delete by ID (example):

wp --path="/Users/cosmodrome/Local Sites/employment-law-aid/app/public" --no-color post delete <ID> --force

8) Final checks

• Spot-check a few published pages in the browser for layout and content.
• Verify taxonomy levels (topic vs. l3-spoke) appear correctly in any filters/menus.

Troubleshooting

• Page exists but content is blank:
  • Ensure the content section after the first # H1 is non-empty.
  • Re-run the sync script; check for WP-CLI errors.
• “No matching WP page” during sync:
  • Run the page creation script (Step 3) and try again.
• Obsidian links published as .md files:
  • Run the Link conversion step (Step 4), then re-sync.

Change log

• 2026-11-10: Added hierarchical page creation and taxonomy assignment to create_missing_pages.py; documented link conversion + validation workflow.