ChatGPT AI Guide: Audit Dandiya Poster Copy and Asset Claims

By TryBuy Editorial

Use ChatGPT to organise a dandiya poster’s approved copy and image notes into a release-review packet. Keep event facts, garment claims and graphic-layout preferences separate. The output should identify missing approvals and contradictions, not invent a venue, date, sponsor or product feature to make the poster feel complete. It is a draft review aid, not permission to publish.

Prepare the approved-copy register

Supply each confirmed event field with its source and exact wording: name, date, location and any other information genuinely required. Mark missing details unresolved. Add who can approve wording where that role has been assigned. Do not ask the model to infer an organiser or decide that a fictional concept is a current event announcement.

Provide image notes separately, including the authorised adult and garment references, the production type and the intended layout role. Keep the supplied product features and inclusions distinct from what the image merely suggests. A kurta shown with accessories does not make those accessories part of the listing, and an AI-generated scene is not evidence that a real event occurred.

Copy the release-review prompt

Audit my approved event-copy register and poster asset notes. Return locked wording, source, unresolved approval, possible contradiction and next responsible role where supplied. Keep event facts, product claims, image issues and layout preferences in separate sections. Do not invent dates, venues, ticket information, sponsors, garment features or permissions. Draft a clean copy handoff only from approved fields and mark missing information clearly. Do not send, publish or label the poster approved on my behalf.

Ask for exact wording to be preserved in the locked-copy column. A stylistic rewrite can be reviewed separately, but it should not silently alter an approved date, place name or instruction.

Filled fictional example

Imagine a fictional poster record has a confirmed title and date but no venue approval. The image note also reports that a headline overlaps the kurta’s lower detail. The packet should create two separate actions: confirm the venue wording and adjust or reconsider the layout. It should not solve the first by inventing a location or the second by moving the garment’s border.

If a generated image contains unapproved lettering, flag it for the appropriate production process rather than treating it as an official source. The example is a planning scenario, not a report of a real event asset or a tested model capability. Keeping the categories separate makes the next human action clearer.

Use controlled revision prompts

• “Show every event fact that lacks an approved source and remove it from the clean-copy draft.”
• “Preserve the locked names and dates exactly while suggesting an optional shorter headline separately.”
• “Separate the image correction from the text-layout change and retain the garment reference IDs.”

When an organiser supplies a new approval, add it to the register before revising the handoff. Do not ask the model to infer approval from a friendly comment or a file labelled final.

Review before any release decision

Compare the draft copy with the approved register, then inspect the actual image and layout. Confirm that product cards point to the intended real item and that styling suggestions remain separate. Check the destination’s current export requirements through an appropriate source. A text packet cannot verify rights, actual garment fidelity or a successful final export without the relevant review.

Use the original practical guide for the practical three-part poster brief and the men’s kurtas collection for exact garment context. Save the register and review actions with the selected asset version. The deliverable should make factual and creative decisions inspectable while leaving publication, permissions and event truth with the people and sources authorised to establish them.

For general instructions and follow-up techniques, see official OpenAI prompting guidance.