How to Sketch a DaVinci Resolve Color Node Tree from a Look Brief
Color trees invent celebrity LUTs. Nodes from the look brief only. Not a product-ad shot list.
The matching generator is the DaVinci Resolve Color Node Tree from a Look Brief (No Celebrity Stills) (Davinci Resolve Color Node Tree From A Look Brief No Celebrity Stills) 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).
Lock Resolve version or write UNKNOWN
Sketch a DaVinci Resolve color-page node tree from a look brief (contrast, sat, grain). No celebrity names or stills, no logos. Not a product-ad video shot list and not a LUT marketplace listing. 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:
Look brief (contrast, sat, grain, palette notes): [Brief]
Resolve version I lock (or UNKNOWN): [Version]
Node count max: [N or 8]
Words I must not use: [Banned]
What I must never invent: [Never]
Output format: [Format]
Language: [Lang]
Timeline color space I allow (or UNKNOWN): [CS]
Reference still policy (NONE / shapes only): [Still]
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.
Serial nodes from the look brief
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: Brief nouns, Version, N, CS, Still, Lang. Forbidden: celebs, logos, LUT SKUs.
- Version lock line; if UNKNOWN write VERSION UNKNOWN.
- Serial spine of up to N nodes with purpose from Brief only.
- Parallel or layer nodes only if Brief asks; else PARALLEL NONE.
- Grain/sat/contrast knobs as qualitative ranges from Brief; numeric values NOT IN INPUTS if absent.
- Still policy: if NONE, write NO REFERENCE STILLS; never celebrity frames.
- Refuse: product-ad shot lists, LUT marketplace copy, inventing celebrity names.
- Compliance pass: Banned and Never. Format as Format. Gaps list of five.
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 still references NONE
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:
- Resolve color node tree from Brief only. No celebrity stills or logos.
- Never invent LUT product SKUs or celebrity frames.
- Lock Version and CS only as pasted.
- 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 celebrity names and logos
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 (davinci resolve color nodes, resolve node tree from look, no celebrity stills) 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.