Update to Magister Marketing

{
  "name": "magister-marketing-plan",
  "description": "Work a brand's Magister marketing plan — find the highest-priority task, read its brief, record progress, add or edit tasks, and change roadmap strategy. Use when the user asks what to work on next, what their marketing plan says, how a task is going, or wants to add, update, complete, or reprioritize marketing work.",
  "included_files": [],
  "skill_md_contents": "---\nname: magister-marketing-plan\ndescription: Work a brand's Magister marketing plan — find the highest-priority task, read its brief, record progress, add or edit tasks, and change roadmap strategy. Use when the user asks what to work on next, what their marketing plan says, how a task is going, or wants to add, update, complete, or reprioritize marketing work.\n---\n\n# Working a Magister marketing plan\n\nA Magister **project** is one brand or client. Each project has one live\n**marketing plan** built from that brand's audit evidence, organized into moves\nacross acquisition, activation, retention, referral, and revenue, and broken\ninto **plan items** — the individual pieces of work.\n\n## Always resolve the project first\n\nEvery plan tool needs a `project_id`. Get it from `list_projects`. Never guess\none, never reuse an id from an earlier conversation, and never construct one.\n\nIf `list_projects` returns exactly one project, use it without asking. If it\nreturns several, ask which brand the user means before doing anything else.\n\n## Answering \"what should I work on?\"\n\nCall `get_next_plan_item`. It returns the single most actionable item **plus a\nself-contained `brief_markdown`** that already explains the work, the reasoning,\nand the acceptance criteria.\n\nPresent that brief. Do not re-derive the task from the plan, do not summarize it\ninto something shorter unless asked, and do not invent steps it does not\ncontain. If the user wants the wider picture instead, `get_marketing_plan`\nreturns the whole plan with its sections and items, and `get_plan_item` returns\none item in full including its recent timeline.\n\n## Recording progress\n\n- `update_plan_item_status` — mark an item running, completed, reopened,\n  blocked, or skipped. Status changes are reversible, so no confirmation is\n  needed beyond the user's request.\n- `add_plan_item_note` — append a progress note to an item's timeline. Use this\n  for findings and context that should survive the conversation.\n\nRecord status honestly. If work partly succeeded, say so in a note rather than\nmarking the item completed.\n\n## Changing the plan itself\n\n- `create_plan_item` adds a new item; `update_plan_item` edits an existing one's\n  title, description, task instruction, expected impact, phase, due date, or\n  priority. Both are ordinary edits.\n- `set_plan_preferences` and `recompile_marketing_plan` are different in kind.\n  **A recompile replaces the live plan's sections and items with a newly\n  generated version.** Any hand-authored item the user added can disappear from\n  the active plan.\n\nBefore calling either one, tell the user plainly that the current plan will be\nregenerated and wait for them to confirm. `recompile_marketing_plan` also\n**spends the organization's credits**, so say that too.\n\nChanging only notes or non-roadmap preferences does not trigger a recompile.\nWhen `set_plan_preferences` does start one, it returns a compile run — poll it\nwith `get_compile_run` rather than calling recompile again.\n\n## Background work\n\nA recompile runs in the background. Poll `get_compile_run` until it reports a\nterminal state. Never start a second recompile because the first has not\nfinished, and never report a new plan before the run completes.\n\n## Useful context tools\n\n- `get_brand` — the audited brand profile: positioning, audience, voice, and\n  messaging pillars. Read this before drafting anything in the brand's voice.\n- `get_audit_health` — the overall health score and per-channel breakdown from\n  the last completed audit. It needs the audit plan id that `get_marketing_plan`\n  supplies.\n- `list_recent_activity` and `list_agent_artifacts` — what has recently been\n  worked on and shipped for the project.\n\n## Do not infer\n\n- Never state a score, status, due date, or priority the tools did not return.\n- Never claim an item is done because you drafted something; only a status write\n  makes that true.\n- If a tool reports the organization is out of credits, tell the user they need\n  to upgrade and stop. Do not retry it and do not substitute another metered\n  tool.\n"
}