How to Build a Processor Addendum Checklist from a Facts Pack
Addendum checklists invent ISO 27001 and SCC Module 2. The pack had a workstation store. Write NO_DATA.
The matching generator is the Processor Addendum Checklist from a Facts Pack (Not Legal Advice, NO_DATA Gaps) (Processor Addendum Checklist From Facts Pack Not Legal Advice No Data) 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).
Print the not-legal-advice banner first
Build a processor-addendum checklist from a facts pack. Mark NO_DATA gaps. Not legal advice. 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:
Facts pack (services, data types, locations I allow): [Facts]
Topics I must score: [Topics]
Processor name I lock (or UNKNOWN): [Processor]
Words I must not use: [Banned]
What I must never invent (ISO, SCCs, sub-processors): [Never]
Output format: [Format]
Language: [Lang]
Max topics: [Max]
Not-advice banner I require: [Banner]
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.
Score each topic YES NO or NO_DATA
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:
- Honesty ledger: Facts nouns, Topics, Processor, Lang, Max. Forbidden: invented ISO, invented SCCs, invented sub-processors.
- Banner: print Banner. Not legal advice. Not a signed addendum.
- Matrix: each Topics line YES, NO, or NO_DATA with a quote or gap. Stay at or under Max.
- Processor: quote Processor or write PROCESSOR UNKNOWN.
- Refuse list: SCC module numbers not in Facts, ISO 27001 certificates, extra sub-processor logos, counsel opinions.
- Sub-processors: only names in Facts. If none, write SUB-PROCESSORS NO_DATA.
- Never: do not write 'we are GDPR certified'.
- Compliance pass: quote Banned and Never hits. Cut them. Count topics vs Max. 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.
Leave sub-processors NO_DATA if unnamed
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:
- Processor addendum checklist from Facts. Not legal advice and not a signed DPA.
- Use NO_DATA rather than inventing controls.
- Stay at or under Max topics.
- 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.
Refuse GDPR certified claims
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 (processor addendum checklist, not legal advice, no_data gaps dpa) 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.