Skip to main content
Emit an operational recommendation (config / setup change): records the action and adds a recommendation card to the Action Engine for review. Creates a to-do for an operational change. Accepting it records the decision; the change itself is made by a person. Common subtypes:
  • ‘enable_model’ — recommend enabling an additional AI engine (perplexity, gemini, etc.) for scans
  • ‘add_persona’ — recommend adding a new persona to the matrix
  • ‘add_competitor’ — recommend a competitor to add to your tracked competitors
  • ‘calibrate_persona’ — recommend a persona-calibration tune for better intent fit
  • ‘expand_journey’ — recommend adding a journey stage to scans
  • ‘add_brand_account’ — recommend a channel account to connect
Use this when:
  • get_brand_context shows a gap in coverage (e.g. only ChatGPT enabled when competitors are also winning on Perplexity).
  • The user asks “what should I set up next?”
Skip when:
  • The fix is structural / on-page (use emit_technical).
  • The fix is content creation (use emit_content).
Inputs:
  • subtype (required): one of the categories above (free-form string ≤ 40 chars).
  • title (required, ≤ 200 chars): one-line summary.
  • description (required): array of step-by-step instructions.
  • effort_minutes (optional, default 15): estimated effort.
  • priority (optional, default ‘medium’).
  • config_target (optional): JSON describing the proposed setting change. Recorded on the action so the user can later apply it.
Response: { actionId, taskId, deepLinkUrl }.

​
Inputs