ads-machine: Command for Claude Code

Ad Intelligence Report

You generate a formatted competitor intelligence report from the Ad Swipe File. This is a snapshot of what's happening in the market right now -- who's running what, what's working, what's dying, and what hooks are winning.

What you produce: A markdown report ready to share with clients, your team, or stakeholders.

---

Config

Read from CLAUDE.md:

Airtable Base ID: YOUR_AIRTABLE_BASE_ID
Ad Swipe File Table: YOUR_SWIPE_FILE_TABLE_ID
Competitors Table: YOUR_COMPETITORS_TABLE_ID
Business: YOUR_BUSINESS_NAME

---

Step 1: Pull the Data

Fetch all ads from the Swipe File:

Use Airtable MCP: list_records
  base_id: {from CLAUDE.md}
  table_id: {Swipe File table ID}
  fields: Ad Archive ID, Competitor, Page Name, Angle Category, Ad Format Type, Display Format, Days Active, Longevity Tier, Start Date, End Date, Hook Copy, Body Text, CTA Type, Is Active, Scrape Date

Also fetch competitor list:

Use Airtable MCP: list_records
  table_id: {Competitors table ID}
  filter: {Status}='Active'
  fields: Name, Niche Tier

---

Step 2: Calculate the Numbers

From the data, calculate:

• Total ads tracked across all competitors
• New ads this week (Start Date within last 7 days)
• Killed ads this week (End Date within last 7 days, no longer active)
• Long-Runners (60d+) -- total count and any new ones
• By competitor: ad count, format breakdown, most common angle
• By angle: which angles have the most Long-Runners
• By format: which formats have the most Long-Runners
• Top 5 hooks from Long-Runners (sorted by Days Active)

---

Step 3: Generate the Report

# Competitor Ad Intelligence Report
**{Business Name}** | Week of {date range} | Generated {today}

---

## Market Snapshot

| Metric | Count |
|---|---|
| Competitors tracked | {N} |
| Total ads in swipe file | {N} |
| New ads this week | {N} |
| Ads killed this week | {N} |
| Long-Runners (60d+) | {N} |

---

## What's Working (Long-Runners)

These ads have been running 60+ days. Someone is paying to keep them alive.

{For each Long-Runner, show:}
1. **{Competitor}** -- {Days Active}d -- {Angle} -- {Format}
   Hook: "{first line of copy}"
   [View in Ad Library]({ad library url})

---

## What's New This Week

{N} new ads launched across {N} competitors.

| Competitor | New Ads | Formats | Dominant Angle |
|---|---|---|---|
| {name} | {count} | {formats} | {angle} |

---

## What Got Killed

{N} ads stopped running this week.

| Competitor | Killed | Avg Days Active | Common Angle |
|---|---|---|---|
| {name} | {count} | {avg days} | {angle} |

Short-lived ads (<7 days) suggest failed tests. Track what they tried and why it might have failed.

---

## Angle Breakdown

| Angle | Total Ads | Long-Runners | Win Rate |
|---|---|---|---|
| {angle} | {count} | {lr count} | {lr/total %} |

Win rate = Long-Runners / Total. Higher win rate means this angle consistently works in your market.

---

## Top 5 Proven Hooks

These hooks have the longest run times in your swipe file.

1. **{days}d** | {competitor} | {angle}
   "{hook text}"

2. **{days}d** | {competitor} | {angle}
   "{hook text}"

3. ...

---

## Recommendations

{Write 3-5 specific, actionable recommendations based on the data above. Examples:}

- "Social proof is the dominant winning angle (X% win rate). Prioritize testimonial-style ads."
- "{Competitor} launched 8 new ads this week, all video UGC. They're testing at volume -- watch for which ones survive to 14d+."
- "No competitors are using [angle]. This is a gap you could test."
- "The top 3 hooks all use question openers. Test question hooks in your next batch."