export declare const RAPID_CHECK = "# /rapid-check \u2014 Rapid Test Daily Check\n\n## What This Does\n\nDaily check during an active rapid test. Paste your numbers in any format, get a verdict. At the end of the test window, produces a final GRADUATE / CONTINUE / KILL decision.\n\n## Trigger Context\nDuring an active rapid test window. User runs this daily (or every other day) after deploying ads. Accepts an optional market slug: `/rapid-check [market-slug]`.\n\n## Execution Protocol\n\n### Step 1: Identify Which Test\n\nCheck pipeline-state.json `rapid_tests` section.\n\n- If the user specified a slug (e.g., `/rapid-check meal-prep-truck-drivers`), use that test.\n- If only one test has `status: \"active\"` or `status: \"ready_to_deploy\"`, use it automatically.\n- If multiple active tests exist, present a numbered list and ask which one:\n  ```\n  Multiple rapid tests active:\n    1. meal-prep-truck-drivers \u2014 Day 2 of 5 | $30/$75 spent\n    2. anxiety-new-moms \u2014 Day 1 of 5 | $15/$75 spent\n  Which test? (number or name)\n  ```\n\n### Step 2: First Check \u2014 Mark as Active\n\nIf this is the first `/rapid-check` for this test:\n- Update `rapid_tests.[slug].status` to `\"active\"`\n- Update `rapid_tests.[slug].deployment.campaign_live` to `true`\n- Update `rapid_tests.[slug].deployment.start_date` to today\n- Calculate and set `rapid_tests.[slug].deployment.end_date` based on duration_days\n- Ask: \"What's your landing page URL?\" and save to `rapid_tests.[slug].deployment.landing_page_url`\n\n### Step 3: Accept Data Input\n\nAccept data in ANY format:\n- \"Spent $14, got 203 impressions, 12 clicks, 1 signup\"\n- Pasted from Meta Ads Manager (table format, screenshot description, CSV)\n- \"12 clicks, 1 signup, about $15 spent\"\n- \"nothing happened yet\" (record as zeros)\n- \"same as yesterday basically\" (duplicate previous day's rough numbers, note uncertainty)\n\nParse and extract: **spend**, **impressions**, **clicks**, **conversions/signups**. If the user omits a metric, calculate from what's available (e.g., CTR from clicks/impressions) or ask only if critical.\n\n### Step 4: Load Context\n\nFrom pipeline-state.json for this rapid test, load:\n- Test parameters (budget, duration, offer type, price point)\n- Decision thresholds (graduate/continue/kill criteria)\n- All previous daily checks\n- Days remaining in window\n- Aggregate metrics so far\n\n### Step 5: Update Aggregate Metrics\n\nAdd today's numbers to cumulative totals:\n- `total_spend += today_spend`\n- `total_impressions += today_impressions`\n- `total_clicks += today_clicks`\n- `total_conversions += today_conversions`\n- Recalculate: `ctr`, `cpc`, `cpa`, `conversion_rate`\n\n### Step 6: Produce Verdict\n\n**Day 1-2 (Insufficient Data):**\n```\nRAPID TEST: [Idea Name] \u2014 Day [X] of [Y]\nSpend: $[today] ($[cumulative] of $[budget])\nData: [impressions] impressions | [clicks] clicks | [conversions] conversions\nCTR: [X]% | CPC: $[X]\n\nStatus: COLLECTING DATA \u2014 Normal for Day 1-2. Need 500+ impressions for meaningful signal.\nAction: Keep running. Check back tomorrow.\nDays remaining: [Y-X]\n```\n\n**Day 2+ with data \u2014 produce ONE of three verdicts:**\n\n**SIGNAL DETECTED:**\n```\nRAPID TEST: [Idea Name] \u2014 Day [X] of [Y]\nSpend: $[cumulative] of $[budget] | [days remaining] days left\nImpressions: [X] | Clicks: [X] | Conversions: [X]\nCTR: [X]% | CPC: $[X] | Cost/conversion: $[X]\n\nSIGNAL DETECTED\n  [The specific metric exceeding threshold]: [value] vs [threshold needed]\n  [If multiple metrics are strong, list the top 2]\nAction: Keep running through Day [Y] to confirm signal strength.\n[If already at graduation threshold: \"Looking strong for graduation. Finish the window for confidence.\"]\n```\n\n**WATCHING:**\n```\nRAPID TEST: [Idea Name] \u2014 Day [X] of [Y]\nSpend: $[cumulative] of $[budget] | [days remaining] days left\nImpressions: [X] | Clicks: [X] | Conversions: [X]\nCTR: [X]% | CPC: $[X] | Cost/conversion: $[X]\n\nWATCHING \u2014 Mixed signal.\n  Working: [what's performing]\n  Not working: [what's underperforming]\nAction: [One specific thing \u2014 \"hold steady\" or \"if you can, swap the headline on your landing page to test\" \u2014 never recommend overhauling the entire test mid-run]\n```\n\n**NO SIGNAL:**\n```\nRAPID TEST: [Idea Name] \u2014 Day [X] of [Y]\nSpend: $[cumulative] of $[budget] | [days remaining] days left\nImpressions: [X] | Clicks: [X] | Conversions: [X]\nCTR: [X]% | CPC: $[X] | Cost/conversion: $[X]\n\nNO SIGNAL \u2014 [specific evidence]\n  [Key metric]: [value] vs [minimum needed]: [threshold]\nAction: [If early and budget remains: \"Give it one more day \u2014 some campaigns take 48h to optimize.\"\n         If Day 3+ with zero signal: \"Consider early kill. You've spent $[X] with [evidence of no demand]. Kill now to save $[remaining budget] for your next test. Run /rapid-check with 'kill' to end this test.\"]\n```\n\n### Step 7: End-of-Window Decision (Final Day or Budget Exhausted)\n\nOn the last day of the test window, OR when budget is exhausted, OR when the user says \"final numbers\" or \"that's it\", produce a FINAL VERDICT:\n\n**GRADUATE:**\n```\nFINAL VERDICT: GRADUATE TO FULL PIPELINE\n\n[Idea Name] showed real market signal:\n  - [Key metric 1]: [value] (threshold was [X])\n  - [Key metric 2]: [value]\n  - Best performing ad: Ad [N] ([angle name]) \u2014 [CTR]% CTR\n  - [Any qualitative signal: replies, comments, shares]\n\nTotal invested: $[total_spend] over [days] days\n\nWHAT HAPPENS NEXT:\n  Run /rapid-graduate [market-slug] to:\n  1. Credit your rapid test research toward the full pipeline\n  2. Carry forward: buyer snapshot, winning ad angle, conversion data\n  3. Begin expanded research \u2192 full pipeline build\n\n  This idea earned deeper investment. The rapid test was your scout.\n```\n\n**CONTINUE (ambiguous signal):**\n```\nFINAL VERDICT: CONTINUE \u2014 Signal is ambiguous\n\n[Idea Name] showed mixed results:\n  - What worked: [specific metrics/elements]\n  - What didn't: [specific metrics/elements]\n  - Best ad: Ad [N] ([angle]) \u2014 worst ad: Ad [N] ([angle])\n\nTotal invested: $[total_spend] over [days] days\n\nOPTIONS:\n  A. Extend 3 more days ($[daily x 3]) \u2014 keep same setup, gather more data\n  B. Adjust and retest \u2014 change [the specific weakest element] and run 3 more days\n  C. Kill \u2014 signal too weak to justify more spend\n\nYour call. Reply with A, B, or C.\n```\n\n**KILL:**\n```\nFINAL VERDICT: KILL\n\n[Idea Name] did not show viable market signal:\n  - [Key evidence 1: e.g., \"CTR 0.4% across all 3 ads \u2014 below 0.8% minimum\"]\n  - [Key evidence 2: e.g., \"$52 spent, 0 email signups\"]\n\nTotal invested: $[total_spend] over [days] days\nLearning: [One sentence \u2014 what this test revealed. E.g., \"This market may not be reachable via Meta ads\" or \"Pain point was real but the promise didn't resonate\"]\n\nThis idea is archived. The learning is saved.\nRun /rapid-test [your next idea] to test another one.\nRun /rapid-status to see your full test portfolio.\n```\n\n### Step 8: Update Pipeline State\n\nAppend daily check entry to `rapid_tests.[slug].daily_checks`:\n```json\n{\n  \"date\": \"[today]\",\n  \"day_number\": [X],\n  \"spend\": [X],\n  \"impressions\": [X],\n  \"clicks\": [X],\n  \"conversions\": [X],\n  \"ctr\": [calculated],\n  \"cpc\": [calculated],\n  \"cost_per_conversion\": [calculated or null],\n  \"verdict\": \"[collecting_data|signal|watching|no_signal|graduate|continue|kill]\",\n  \"notes\": \"[any user-provided context or observations]\"\n}\n```\n\nUpdate `rapid_tests.[slug].aggregate_metrics` with new cumulative values.\n\nOn final verdict:\n- Update `rapid_tests.[slug].status` to `\"graduated\"`, `\"extended\"`, or `\"killed\"`\n- Update `rapid_tests.[slug].verdict` to `\"graduate\"`, `\"continue\"`, or `\"kill\"`\n- For GRADUATE and KILL: auto-trigger `/lessons capture` with the test data (what worked, what didn't, market characteristics)\n\nOn CONTINUE option A (extend):\n- Add `duration_days` to the existing duration\n- Update `end_date`\n- Set status to `\"active\"` (stays active)\n\nOn CONTINUE option C (user kills):\n- Process as KILL verdict above\n\n### Step 9: Early Kill Handling\n\nIf the user says \"kill\" or \"kill it\" or \"stop this test\" at any point (not just end of window):\n- Calculate remaining budget saved: $[budget - total_spend]\n- Process as KILL verdict\n- Present: \"Killed [idea name]. Saved $[remaining]. Run /rapid-test with your next idea.\"\n\n## Rules\n- At $15/day, daily fluctuations ARE the data. Don't wait for 7-day trends like the full pipeline does at $50/day. Focus on absolute thresholds, not trajectory.\n- NEVER recommend rebuilding the landing page or overhauling the campaign mid-test. The test is 3-5 days. Either signal exists or it doesn't. The only mid-test adjustment allowed is pausing the worst-performing ad to concentrate budget on the others.\n- If user hasn't run /rapid-check and the test window has passed, flag it: \"Your [idea name] test window ended on [date]. Run /rapid-check with your final numbers to get a verdict.\"\n- After a KILL verdict, always suggest the next action: \"Run /rapid-test [idea] to test your next one.\"\n- After a GRADUATE verdict, always suggest: \"Run /rapid-graduate [market-slug] to enter the full pipeline.\"\n\n## Autonomy\nTier 1 \u2014 Full Auto for daily check analysis and state updates.\nTier 2 \u2014 Notify & Proceed for GRADUATE verdict.\nTier 3 \u2014 Recommend & Wait for CONTINUE verdict (user chooses path).\nTier 4 \u2014 Escalate for KILL only if user has killed 3+ consecutive ideas with zero conversions across all of them (pattern suggests a deeper issue \u2014 recommend stepping back to do one full /scout to recalibrate).\n\n## Output\nInline verdict. State updated in pipeline-state.json. No separate file created \u2014 daily checks are tracked in pipeline state.\n\n## Next Step\n- Daily during window \u2192 `/rapid-check` again\n- GRADUATE \u2192 `/rapid-graduate [market-slug]`\n- CONTINUE \u2192 extend, adjust, or kill per user choice\n- KILL \u2192 `/rapid-test [next idea]` or `/rapid-status` to review portfolio";
//# sourceMappingURL=rapid-check.d.ts.map