What problem does it solve? Keyword research through the AIOS platform involves a paid provider call that can be silently blocked by a cost gate, so a naive run can return an empty bank that looks like a valid result. This Skill queues the research job, verifies what actually landed by comparing the keyword bank before and after, and reports only fields the backend truly returns. ## Core Features & Use Cases - Seed-based keyword research: Queue a research job for a client from a seed term and verify the saved-keyword count actually changed before reporting success. - Bank, cluster, and conflict reporting: Read saved keywords with volume, raw provider difficulty, CPC, intent, and opportunity scores, plus cluster pillar maps and cannibalization conflicts where one URL targets keywords with different intents. - Honest failure handling: Detect silent cost-gate blocks and dormant providers instead of presenting an empty bank as "no keywords found", and never re-queue a blocked run since that is a budget decision. - Use Case: An SEO operator asks for keyword ideas around "local plumber seo" for a client; the Skill resolves the client, queues the research, confirms 40 new keywords landed in the bank, and renders the cluster map and cannibalization conflicts. ## Quick Start Ask the assistant to research keywords for a client from a seed term, for example: run keyword research for client Acme with the seed term "emergency plumbing services" and show the clusters and cannibalization conflicts.
