← ClaudeAtlas

ads-hintslisted

Develop, group and prioritise context hints for OpenAI Ads, and derive account-wide negative keywords. Use this skill for topic research, context hints, ad targeting, exclusions, negative keywords, or when deciding which conversations an ad should appear in.
devaimani/openai-ads-mcp · ★ 0 · AI & Automation · score 76
Install: claude install-skill devaimani/openai-ads-mcp
# Context hints ## What they are Free text describing when an ad is relevant. The platform evaluates it semantically against live conversations. What they are **not**: keywords. There are no match types, no per-hint bid and no per-hint reporting. > "Provide a list of descriptions or keywords for when the product or service > might be useful to show." ## How many The API accepts up to 2000 per ad group. That is a technical ceiling, not a recommendation. **Because there is no per-hint reporting, the ad group is the smallest unit of measurement.** 200 hints in one group means seeing the outcome but never the cause. A success cannot be repeated and a failure cannot be isolated. **Guideline: 10 to 25 hints per group, tightly scoped.** Five groups of 15 beat one group of 75. A hint belongs in a group only if it serves the same intent as the rest. The moment you hesitate, it belongs in its own group. ## Phrasing Hints are free text read semantically. Use natural language, not keyword syntax. | Instead of | Better | |---|---| | `process automation smb` | `A mid-sized company is looking for ways to automate repetitive workflows` | | `chatbot price` | `Someone is asking what a chatbot for their website costs` | | `gdpr ai` | `The topic is using AI in a company in a data-protection-compliant way` | Descriptions of the conversational situation outperform search terms, because matching is semantic and questions put to a chat assistant are longer and more discursive than searc