← CaptinoCONTENT HISTORY

Update to Captino

Snapshot Nov 9, 2026 · 18:08 UTC · version 0.4.1

Collection source: downloaded plugin package.

WHAT CHANGED

First saved snapshot

No earlier snapshot is available to establish a change.

Save Captino and check future changes in Following.

Up to 3 plugins free · no card required.

Compare saved observations

Download comparison JSON
Full technical diff · 0 changed fields
Full snapshot data
{
  "description": "Analyze how a Captino funnel performs - visits, unique visits, opt-ins and opt-in rate over a period, the change against the previous period, the ranking of all funnels, or a head-to-head of two funnels. Use whenever the user asks \"how is my funnel doing\", \"which funnel performs best\", \"did traffic or opt-ins go up\", \"compare funnel A with B\", \"funnel stats for last week\", or any question about funnel numbers, even if they don't say \"performance\".",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 480
    }
  ],
  "name": "funnel-performance",
  "skill_md_contents": "---\nname: funnel-performance\ndescription: Analyze how a Captino funnel performs - visits, unique visits, opt-ins and opt-in rate over a period, the change against the previous period, the ranking of all funnels, or a head-to-head of two funnels. Use whenever the user asks \"how is my funnel doing\", \"which funnel performs best\", \"did traffic or opt-ins go up\", \"compare funnel A with B\", \"funnel stats for last week\", or any question about funnel numbers, even if they don't say \"performance\".\n---\n\n# Funnel performance\n\nAnswers questions about funnel traffic and opt-ins using the Captino MCP server.\n\n## Prerequisites\n\nUses these Captino MCP tools: `list_funnels`, `get_funnel`, `list_funnel_performance`,\n`get_funnel_performance`, `compare_funnel_performance_with_previous_period`,\n`get_funnel_step_performance`, `compare_funnels`. They need **Funnel: read** and\n**Performance: read** access, granted when connecting Captino. If the tools are\nmissing or a call returns 401/403, run the `captino-setup` skill instead of\nguessing. Never fabricate numbers; if a tool is unavailable, say so.\n\n## Workflow\n\n### 1. Resolve the funnel\n\n- Ids are GUIDs. Never guess one. If the user names a funnel, call `list_funnels`\n  with `query` set to the name and pick the match. If several match, list them\n  (name, domain, active) and ask which one.\n- If the user says \"my funnels\" or \"all funnels\", skip to ranking (step 3a).\n- Keep the `id`, `name`, `isActive` and `visitUrl` from the result. A funnel with\n  `isActive: false` or `visitUrl: null` cannot receive traffic; mention this if its\n  numbers are zero.\n\n### 2. Resolve the period\n\nTools take `from` and `to` as ISO 8601 with offset (for example\n`2026-09-01T00:00:00+02:00`). Work out the dates from today's date in the\nuser's timezone; the server converts to UTC.\n\n| User says | Period |\n| :-- | :-- |\n| nothing / \"recently\" | last 30 days, ending today |\n| \"this week\" | Monday of this week to today |\n| \"last week\" | previous Monday to previous Sunday |\n| \"this month\" / \"last month\" | first to last day of that month |\n| \"last 7 days\", \"since launch\" | as stated; use `createdAt` from `list_funnels` for launch |\n\nAlways state the period you used in the answer.\n\n### 3. Pick the tool for the question\n\n| Question | Tool | Notes |\n| :-- | :-- | :-- |\n| a. Which funnel is best / rank my funnels | `list_funnel_performance` with `rankBy` | `OptIns` for \"which converts best\", `UniqueVisits` for reach, `AllVisits` (default) for raw traffic. All-time only, returns names and values, no ids. |\n| b. How did funnel X do in a period | `get_funnel_performance` | Returns `dates`, `lines` (one series per metric), `totalVisitors`, `totalOptIns`, averages per day and week. |\n| c. Is it up or down | `compare_funnel_performance_with_previous_period` | `selectedGraphData` is the requested period, `pastGraphData` the equally long period before it. |\n| d. Traffic per step over time | `get_funnel_step_performance` | One line per step. For drop-off and conversion per step use the `funnel-dropoff` skill. |\n| e. Funnel A vs funnel B | `compare_funnels` | All-time `allVisits` and opt-ins for `leftFunnel` and `rightFunnel`. |\n\nFor an overview of one funnel with no specific question, call b and c together\nfor the default period, plus `list_funnels` for the all-time counters.\n\n### 4. Interpret\n\n- Opt-in rate = `totalOptIns / totalVisitors`. Say \"no traffic\" rather than\n  dividing by zero.\n- Change vs previous period = `(current - past) / past`. If `past` is zero, report\n  the absolute numbers instead of a percentage.\n- Use the daily `lines` to spot spikes or dead days; name the date of the best\n  and worst day when it is useful.\n- All-time counters on `list_funnels` (`uniqueVisitors`, `pageviews`, `optIns`,\n  `formResponses`) are not comparable to period numbers; don't mix them in one\n  table without labelling.\n\n### 5. Report\n\nLead with the answer in one or two sentences, then one compact table. Example:\n\n> Spring Campaign got fewer visits than the week before but converted better.\n>\n> | Metric (1-7 Sep vs 25-31 Aug) | This period | Previous | Change |\n> | :-- | --: | --: | --: |\n> | Visits | 412 | 530 | -22% |\n> | Opt-ins | 38 | 31 | +23% |\n> | Opt-in rate | 9.2% | 5.8% | +3.4 pp |\n\nEnd with at most two concrete observations (for example the day traffic dropped,\nor that the funnel is inactive). Offer the `funnel-dropoff` skill when the opt-in\nrate is the problem.\n"
}

SHA-256 of public snapshot: f8bcce30f245918b603b1e67fbfa44826b79dec69278f50db609491e620140aa