Daily Brief

1. Purpose

A single morning read that answers: "what needs my attention today, in priority order?" It doesn't re‑implement detection logic — it orchestrates the skills that already do it, then synthesizes their output into a ranked digest. Run once per day; it's the entry point into the loop.

2. Chain of Skills

Run these in order; keep only the specified items from each output.

| Step | Skill                                     | Keep                                                                                                                          |
| :--- | :---------------------------------------- | :---------------------------------------------------------------------------------------------------------------------------- |
| 1    | [am-i-on-track](am-i-on-track.md)         | 🔴 Critical and 🟡 Warning pacing alerts only (drop 🟢)                                                                       |
| 2    | [whats-running](whats-running.md)         | ⚠️ flagged campaigns only (drop ✅)                                                                                           |
| 3    | [bleeders-winners](bleeders-winners.md)   | Top 3 Bleeders + top 3 Winners by spend impact                                                                                |
| 4    | [fatigue-detector](fatigue-detector.md)   | 🔴 Critical entries only (drop 🟡 Warning)                                                                                    |
| 5    | [copy-rotation](copy-rotation.md)         | Generate **1 replacement variant pack** for every creative flagged to pause in steps 3 (Bleeders) or 4 (Fatigue). Dedup pairs. |

3. Ranking & Dedup Rules

• Section order by spend‑at‑risk: Pacing → Leaks → Stop/Scale → Fatigue → Replacements.
• If the same ad appears in both Bleeders and Fatigue, keep only the Bleeders entry (stronger signal, explicit action) — and dedup the replacement pack for it.
• Collapse empty sections — don't render a header with no content.
• Cap the alert portion at 12 items across sections 1–4; Replacements are not counted against the cap (they're outputs, not alerts) but limit to one pack per paused creative.

4. Executive Summary (top of brief)

A single scannable line. Skip counts that are zero.

"[N] campaign(s) [over|under]pacing · [N] leak(s) · [N] to pause, [N] to scale · [N] creative(s) critical · [N] replacement pack(s) ready."

5. Operational Output: Morning Brief

Daily Brief — Mon, Apr 22

Summary: 2 campaigns overpacing · 1 leak · 3 to pause, 2 to scale · 1 creative critical · 4 replacement packs ready.

Pacing (2)

• 🔴 Retargeting — Conversions: projected $580 vs. $450 budget (+29%)
• 🟡 Summer Launch — Traffic: 18% above pace

Leaks (1)

• ⚠️ Q1 Test Campaign ($50/day) — scheduled to end 3/1, still running

Stop / Scale (5)

• 🩸 3 Bleeders: AD_412 · AD_501 · AD_733 — see bleeders-winners for copy + action
• 🏆 2 Winners: AD_388 · AD_390 — +20% budget suggested (see budget-shifter for source)

Creative Fatigue (1)

• 🔴 AD_238 (Notes App Hero) — CTR dropped 34% in 3 days. Pause and replace with pack below.

Replacement Packs (4) — via copy-rotation

• AD_412 → 3 headlines + 3 bodies (Horsemen: Money · Status · Fear)
• AD_501 → 3 headlines + 3 bodies (Horsemen: Time · Status · Money)
• AD_733 → 3 headlines + 3 bodies (Horsemen: Fear · Money · Time)
• AD_238 → 3 headlines + 3 bodies (Horsemen: Time · Fear · Status)

(Full copy expanded under each pack — headline, body variants with word counts, and assigned Horseman.)