Endnote Link

EndnoteLink Skill

Converts a build_doc.py-generated manuscript .docx to EndNote-ready format. Runs CrossRef metadata validation (retraction screening, author/year checks) before converting. Outputs an _endnote.docx with {Author, Year #N} temp citations and a .ris file for import into EndNote.

Triggers

Invoke when the user says any of:

• "endnote link", "endnote-link"
• "convert to endnote", "export to endnote"
• "endnote import", "prepare for endnote"
• "convert citations for endnote"
• "export citations", "reference manager export"

Prerequisites

• build_doc.py has been run and a .docx exists
• references/cite_map.json exists (auto-generated by build_doc.py)
• references/key_papers.bib is populated (by /literature:card)

If cite_map.json is missing: tell the user to run python3 build_doc.py first — the template writes it automatically on every build.

Phase 1 — Locate project

Find:

1. build_doc.py — establishes project root
2. references/cite_map.json — required
3. references/key_papers.bib — required
4. The .docx to convert — default: most recently modified .docx in project root

# Confirm all three exist
ls {project_dir}/build_doc.py
ls {project_dir}/references/cite_map.json
ls {project_dir}/references/key_papers.bib

# Find the docx
ls -t {project_dir}/*.docx | head -5

Pre-flight report:

Pre-flight check
  build_doc.py        ✓
  cite_map.json       ✓  ({N} references)
  key_papers.bib      ✓  ({N} entries)
  docx to convert     ✓  {filename}

If cite_map.json is absent: stop and tell the user:

cite_map.json not found. Run python3 build_doc.py to regenerate it — the template writes it automatically. Once rebuilt, run /literature:endnote-link again.

Phase 2 — Run conversion

Run the converter from this skill's references folder:

python3 {base_dir}/references/convert.py {docx_path} \
    --project {project_dir} \
    --email fakoredesodiq@gmail.com

Where {base_dir} is the "Base directory for this skill" path shown in this skill's system reminder header (e.g., /home/sodiq/.claude/plugins/cache/my-skills/literature/0.1.0/skills/endnote-link).

The script runs in four numbered steps and prints progress:

[1/4] Loading cite_map + bib...
[2/4] CrossRef validation (author · year · retraction)...
[3/4] Fetching RIS...
[4/4] Converting citations and writing outputs...

Flags available:

• --skip-validation — skip CrossRef validation (faster, no metadata check)
• --dry-run — show DOI lookup preview without writing any files
• --output {path} — custom output docx path
• --ris {path} — custom RIS output path

Phase 3 — Interpret CrossRef validation output

The script prints a validation block before converting:

══════════════════════════════════════════════════════════
  CrossRef Metadata Validation
══════════════════════════════════════════════════════════
  [ 1] hackett-2018                   ✓ clean
  [ 2] skillback-2022                 ✓ clean
  [ 3] collyer-2022                   ✓ clean
  ...
──────────────────────────────────────────────────────────
  Clean: 5   Warnings: 0   Alerts: 0
══════════════════════════════════════════════════════════

Status symbols:

Action on ALERT (retraction):

1. Remove the citation from the manuscript
2. Update the card: set tags: [RETRACTED] in card.md, update meta.json
3. Rerun build_doc.py and then /literature:endnote-link again

Action on Author/Year warning:

1. Open references/key_papers.bib
2. Verify the author and year fields against the paper
3. Correct if needed, then rerun — or accept if the discrepancy is a known CrossRef data issue (epub vs print, hyphenated surnames, etc.)

Phase 4 — Report outputs

After the script finishes, report:

══════════════════════════════════════════════════════════
  EndnoteLink complete
══════════════════════════════════════════════════════════
  CrossRef validation:  {N_ok} clean | {N_warn} warnings | {N_alert} alerts
  RIS entries:          {N} written to {stem}.ris
  Converted docx:       {stem}_endnote.docx

  Next steps in EndNote:
  1. File → New  (fresh library for this paper — do not reuse old ones)
  2. File → Import → File → {stem}.ris
     Import Option: Reference Manager (RIS)
  3. Open {stem}_endnote.docx in Word
  4. Delete the old reference list at the bottom of the document
  5. EndNote ribbon → Update Citations and Bibliography
══════════════════════════════════════════════════════════

What this does NOT do

• Does not push citations into EndNote directly (EndNote has no open API)
• Does not modify build_doc.py or key_papers.bib
• Does not reformat the reference list style (EndNote handles that via its output style)
• Does not run claim-audit — run /literature:claim-audit {project_dir} separately if needed

Quick reference