Ads Quick

60-Second Ad Readiness Snapshot

Portability note: Uses the WebFetch and WebSearch tools (host-agnostic — work in any Claude Code environment, no OmegaOS VPS dependency). No file I/O, no shell, no VPS-only infra. If WebFetch/WebSearch are unavailable in the host, ask the user to paste the homepage content instead of fetching.

Single-voice craft (do NOT parallelize)

This skill is a single-pass speed tool — it is the deliberate anti-fan-out. Its entire value is one fast, coherent verdict in under 60 seconds. Never launch subagents, a Workflow fan-out, a team, or adversarial-verify lanes here — doing so destroys the core capability (speed). One model, one pass, one scorecard. Craft upgrades that keep it sharp:

1. Anchor every score to an observed cue, not a vibe. Each dimension score must cite a concrete element you actually saw in the fetched page (e.g. "CTA 14/20 — only generic 'Contact Us' above the fold, no specific offer button"). No cue observed -> score the absence, never guess upward.
2. One specific next step beats a list. The "NEXT STEP" line must name the single highest-leverage fix (the gap that moves the score the most), phrased as an action the user can do today — not generic advice.
3. Calibrate to the business type before scoring budget/CPC. Identify the business type from the page first; a restaurant and a B2B SaaS read the same homepage cues very differently. Pull CPC/CPA/budget from the matching row, never a blended average.

Skill Purpose

Perform a rapid ad readiness assessment of a website or business without launching subagents. Fetches the homepage, evaluates value proposition strength, offer clarity, CTA quality, landing page readiness, and trust signal presence. Outputs a compact scorecard with a platform recommendation and estimated starting budget. The entire output must be under 40 lines.

When to Use

• User wants a quick gut-check before investing in a full ad strategy
• User asks "should I run ads?" or "am I ready for ads?"
• User wants a fast assessment without waiting for the full 5-agent analysis
• User is exploring whether paid advertising makes sense for their business
• Triggered by /ads quick <url>

When NOT to Use

• User wants a detailed, comprehensive strategy (use /ads strategy instead)
• User already has ad data to analyze (use /ads audit instead)
• User needs creative briefs (use /ads creative instead)
• No URL is provided (this skill requires a URL to fetch)

How to Execute

Step 1: Fetch and Analyze the Homepage

Use WebFetch to retrieve the homepage content. Extract these elements:

WebFetch(url) -> Analyze:
1. Headline / Hero Section — What is the main promise?
2. Value Proposition — Is it clear what they sell and why it matters?
3. CTA(s) — What action are they asking visitors to take?
4. Offer — Is there a specific offer (pricing, free trial, discount)?
5. Trust Signals — Reviews, testimonials, badges, client logos, certifications?
6. Visual Quality — Professional images, consistent branding, modern design?
7. Mobile Indicators — Responsive hints, fast-loading signals?
8. Contact Info — Phone, email, chat, address visible?

Step 2: Score 5 Dimensions

Rate each dimension on a 0-20 scale:

1. Value Proposition Strength (0-20)

2. Offer Clarity (0-20)

3. CTA Quality (0-20)

4. Landing Page Readiness (0-20)

5. Trust Signal Presence (0-20)

Ad Readiness Score = Sum of all 5 dimensions (0-100)

Step 3: Determine Platform Recommendation

Based on the business type and readiness score, recommend the best starting platform:

Secondary Platform Recommendation: After the primary platform is established (30-60 days), recommend a secondary platform:

Step 4: Estimate CPC and CPA Ranges

Provide realistic cost expectations based on industry and platform:

CPC Ranges by Industry and Platform:

Expected CPA by Business Type:

Step 5: Competitive Quick Check

Before finalizing the snapshot, do one quick competitive signal check:

WebSearch query:

"[Industry]" "[City]" ads OR advertising OR sponsored

Note: Are competitors actively advertising? If yes, the market is validated but competitive. If no, there may be an opportunity or the market may not support paid ads.

Step 6: Generate Readiness Verdict

Output Format

Output directly to terminal. Do NOT save a file. Keep output under 40 lines.

============================================
  AD READINESS SNAPSHOT: [Business Name]
  URL: [url]
============================================

  AD READINESS SCORE: [XX]/100 — [Verdict]

  SCORECARD:
  Value Proposition ... [XX]/20 [status word]
  Offer Clarity ....... [XX]/20 [status word]
  CTA Quality ......... [XX]/20 [status word]
  Landing Page ........ [XX]/20 [status word]
  Trust Signals ....... [XX]/20 [status word]

  TOP 3 STRENGTHS:
  1. [strength]
  2. [strength]
  3. [strength]

  TOP 3 GAPS TO FIX:
  1. [gap + specific fix]
  2. [gap + specific fix]
  3. [gap + specific fix]

  RECOMMENDED PLATFORM: [Platform]
  ESTIMATED STARTING BUDGET: $[X]-$[Y]/month
  EXPECTED CPC RANGE: $[X]-$[Y]

  NEXT STEP: [One specific action to take]
  Full strategy: /ads strategy [url]
============================================

Output Contract & Verify

Produces: exactly ONE terminal scorecard in the block above — no file written. The scorecard MUST contain: business name + URL, an Ad Readiness Score XX/100 + matching verdict, all 5 dimension scores XX/20, exactly 3 strengths, exactly 3 gaps (each with a specific fix), a recommended platform, a starting budget range, a CPC range, and one NEXT STEP line ending with Full strategy: /ads strategy [url].

VERIFY before emitting (self-check, single pass):

1. The 5 dimension scores SUM to the headline /100 score (arithmetic check).
2. The headline verdict matches the score band in the Step 6 table.
3. The platform + budget + CPC come from the row matching the detected business type.
4. Output is ≤ 40 lines and no file was created.
5. Every score traces to a cue actually observed in the fetched page (evidence rule below).

Evidence / no-hallucination guardrail:

• Score ONLY from what the fetched homepage actually shows. Do not invent reviews, pricing, or features that were not present in the fetched content.
• If WebFetch fails or returns thin/blocked content, say so explicitly and ask for a different URL or pasted content — never fabricate a scorecard from assumptions. A blocked/empty fetch is an ABORT, not a guessed PASS.
• Be honest: do not inflate scores to be encouraging. An unready site scored as ready wastes the user's ad budget.

Important Rules

• This is a SPEED tool — complete the entire analysis in under 60 seconds
• Do NOT launch any subagents — this is a single-pass analysis
• Output must be under 40 lines — be concise
• Do NOT save a file — output to terminal only
• Always end with a recommendation to run /ads strategy for the full analysis
• If the homepage cannot be fetched, explain the issue and suggest the user provide a different URL
• Be honest about readiness — do not inflate scores to be encouraging
• Budget recommendations should be realistic monthly minimums, not aspirational
• Always consider the business type when recommending platforms
• One specific next step is more valuable than a list of generic advice