Update to Marketing Swarm

{
  "name": "budget-media-allocation",
  "description": "Use when deciding how to distribute paid-media budget across campaigns, audiences, geographies, or platforms, including cross-platform comparisons and attention-cost opportunities.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 285
    }
  ],
  "skill_md_contents": "---\nname: budget-media-allocation\ndescription: Use when deciding how to distribute paid-media budget across campaigns, audiences, geographies, or platforms, including cross-platform comparisons and attention-cost opportunities.\n---\n\n# Budget and Media Allocation\n\nCompare marginal opportunity, not just headline ROAS.\n\n## Required framing\n\nEstablish:\n\n- total budget and time horizon\n- objective and primary KPI\n- constraints and minimum viable spend per campaign/platform\n- current allocations\n- comparable performance windows\n- whether metrics are platform-attributed or experimentally validated\n\n## Analysis\n\n### Normalize the comparison\n\nDo not compare platforms as if attribution, auction dynamics, conversion windows, and funnel roles were identical. Note differences in intent, reach, frequency, conversion lag, and measurement.\n\n### Estimate marginal value\n\nWhere data permits, inspect how performance changes as spend changes. Prefer spend-response evidence over a single average ROAS/CPA.\n\nUse signals such as:\n\n- CPM and qualified reach\n- CTR and click quality\n- CVR and CPA\n- revenue/value and margin-adjusted return\n- saturation/frequency\n- impression share or lost opportunity\n- recent budget-response behavior\n- creative capacity and fatigue\n\nDo not call lower CPM an arbitrage opportunity unless downstream quality makes the inventory economically useful.\n\n### Build allocation scenarios\n\nProvide at least three when the user wants a decision:\n\n- **protect**: prioritize stability and proven efficiency\n- **balanced**: shift limited budget toward stronger marginal opportunities\n- **explore**: reserve controlled spend for uncertain but promising opportunities\n\nFor each scenario state allocation, rationale, main risk, and measurement plan.\n\n### Cross-platform rule\n\nTreat platform-reported conversions as a measurement input, not a common currency. If platform overlap or attribution conflict could reverse the decision, route to `causal-attribution` before recommending a large shift.\n\n## Output contract\n\nReturn:\n\n- current allocation diagnosis\n- opportunities and constraints\n- scenario table\n- recommended scenario with confidence\n- expected directional impact or range, only when data supports it\n- guardrails and rollback/stop conditions\n- measurement needed to learn from the shift\n\nDo not claim a precise future ROAS from sparse historical averages.\n"
}