Update to Magister Marketing

{
  "name": "magister-ai-visibility",
  "description": "Track and report how a brand appears in AI-generated answers — ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — using Magister AI-visibility prompt sets. Use when the user asks about AI visibility, GEO, generative engine optimization, whether AI assistants mention or recommend their brand, or wants to run or read a visibility report.",
  "included_files": [],
  "skill_md_contents": "---\nname: magister-ai-visibility\ndescription: Track and report how a brand appears in AI-generated answers — ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews — using Magister AI-visibility prompt sets. Use when the user asks about AI visibility, GEO, generative engine optimization, whether AI assistants mention or recommend their brand, or wants to run or read a visibility report.\n---\n\n# Magister AI visibility (GEO)\n\nAn **AI-visibility set** is a group of tracked prompts — the questions a brand's\nbuyers actually ask. Magister runs each prompt against several AI answer engines\nand scores how the brand appears. Each execution of a set is a **round**, and\ncomparing rounds over time is the point of the product.\n\nResolve `project_id` from `list_projects` first; never guess one.\n\n## Reading existing visibility\n\n1. `list_ai_visibility_sets` — the project's tracked sets.\n2. `get_ai_visibility_set` — one set's prompts, rounds, and latest scores.\n3. `render_ai_visibility_report` — a readable score summary plus an interactive\n   report with the trend across rounds, scores by AI model, and per-prompt\n   detail.\n\nPrefer `render_ai_visibility_report` when the user wants to *see* how they are\ndoing. Call it once; it needs no polling and performs no network fetches of its\nown.\n\n## Rounds that are not comparable are deliberately excluded\n\nThe trend only charts rounds measured the same way. A round that was scored\nagainst a different set of models, or that failed to measure enough of its\nprompts, is **left out of the trend on purpose** rather than averaged in.\n\nSo the chart can legitimately show fewer points than the raw round count. That\nis correct behavior, not missing data. Do not describe it as an error, and do\nnot try to reconstruct the excluded rounds.\n\n## Running a new round costs credits\n\n`run_ai_visibility_set`, `create_ai_visibility_set`, and\n`suggest_ai_visibility_prompts` all **spend the organization's credits**.\n\nBefore starting one, tell the user it is a metered run. If they only want to\nknow their current standing, read the existing set instead — reads are free.\n`get_credit_balance` reports the balance and whether metered tools are currently\nblocked.\n\n## Rounds are asynchronous\n\n`run_ai_visibility_set` starts a round; it does not return scores. Poll\n`get_ai_visibility_set` until the round reaches a terminal state.\n\nNever start a second round because the first has not finished — that spends\ncredits again for the same answer. Rounds take minutes, not seconds; tell the\nuser it is running rather than waiting silently.\n\n## Creating a new set\n\nUse `suggest_ai_visibility_prompts` to generate site-grounded prompt candidates,\nshow them to the user, and create the set only from prompts they accept.\n`create_ai_visibility_set` can start the set's first round immediately, so say\nthat before calling it.\n\nWrite prompts the way a buyer would ask them — a category question the brand\nshould win, not the brand's own name alone.\n\n## Do not infer\n\n- Never state a visibility score, a change, or a model's answer the tools did\n  not return.\n- Never claim a brand is or is not mentioned by an AI assistant without a round\n  that measured it.\n- If a tool reports the organization is out of credits, say plainly that they\n  need to upgrade, and stop. Do not retry and do not reach for another metered\n  tool.\n"
}