Home/Blog/How To Clean A Zotero Csv Column Header Without Inventing Fields
Blog

How To Clean A Zotero Csv Column Header Without Inventing Fields

P
promptstudio
How To Clean A Zotero Csv Column Header Without Inventing Fields

CSV cleaners invent DOI fills. Keep, rename, or drop; never backfill.

The matching generator is the Zotero CSV Column Cleaner from an Export Header (No Invented Fields) (Zotero Csv Column Cleaner From Export Header No Invented Fields) prompt. Browse related cards in the PromptDig library (Browse more prompts). When a filled run survives, share the version you actually use (Share a prompt).

Map every header name to keep, rename, or drop

Clean a Zotero CSV export header into a keep/rename/drop map. Never invent columns or fill empty cells with guessed DOIs. Start by filling Inputs, not by asking the model to remember last week's run. If a field is blank, write NONE or NOT IN INPUTS and leave it blank through Generate. The card is built so the model cannot honestly invent a number, owner, URL, or command that you did not paste.

Paste these fields before you hit run:

Pasted CSV header (comma-separated names): [Header]
Columns I must keep: [Keep]
Rename map I lock (old=new or NONE): [Renames]
Words I must not use: [Banned]
What I must never invent: [Never]
Output format: [Format]
Language: [Lang]
Max columns discussed: [Max]

That inventory is the honesty ledger. Anything that does not appear there is forbidden in the draft. If you catch yourself adding a nice-to-have after the run, you are no longer using the card. You are ghostwriting. Put the extra fact in Inputs and run again.

Rename only pairs locked in Inputs

Generate is numbered on purpose. Do not skip a step because the first paragraph looked done. The early steps exist to stop later prose from smuggling claims.

Walk the Generate list in order:

  1. Honesty ledger: header count, Keep count, Renames status, Lang, Max. Forbidden: invented columns, invented DOI fills.
  2. Header map: each name keep/rename/drop against Keep and Renames.
  3. Clean spec: up to Max columns. Drop list for header names not in Keep.
  4. Rename table: only pairs in Renames.
  5. Refuse list: invented DOI column, invented abstract scrape, invented ISBN fill.
  6. Diff notes: Banned/Never cuts.
  7. Empty-cell note: never backfill empty DOI/URL/pages cells.
  8. Compliance pass: Banned/Never hits. Column count vs Max. Gaps list of five. Format as Format.

If a step asks for a version lock, quote the version from Inputs in the output. If a step asks for a refuse list, keep the refuse list in the published artifact, not in a sidebar you delete. Reviewers should see what the model was not allowed to do.

Refuse invented ISBN columns

Most failures are the same shape: a missing field gets a confident fill. A conversion rate appears. A Gradle task appears. A flash point appears. A caption appears on a job that asked for slide text only. Your review is to search the draft for numbers, names, and commands, then grep Inputs. No match means cut.

Honor the constraints as hard stops, not vibes:

  • Column cleaner from Header only. Not BibTeX and not a style formatter.
  • Never invent fields or fill empty cells.
  • Stay at or under Max columns discussed.
  • No emojis.

When the card says not legal advice, not certification, not an exam dump, or not a caption engine, that sentence belongs at the top of the output. Deleting it to look more finished is how you inherit risk.

Leave empty DOI cells empty

Finish with the compliance pass the prompt already asks for. Quote the banned-word hits. Cut them. Print character counts when the job has a cap. Print word counts when the job has a budget. List gaps as gaps. Five missing facts are more useful than one smooth paragraph.

Tags on the card (zotero csv cleaner, no invented fields, zotero export header) are a reminder of the job shape, not an invitation to wander into a neighboring cluster. If you need a different surface, open a different PromptDig card rather than stretching this one.

Fill the card, then run

Replace every bracket. Run on ChatGPT, Claude, or Gemini. Read the ledger first, then the artifact. If the model invents a commit, KPI, DOI, PEL, bid, or logo, discard the run. Tighten Inputs. Run again. Share the filled card that survived, not the first draft that sounded done.