agentsclimarketplace

Neural audit

Skill abouchard11/midnight-seo-skills/skills/neural-audit

Full portfolio analysis — TAM, priority matrix, attack sequencing across all live domainsFrom its SKILL.md

Install
npx -y skills add abouchard11/midnight-seo-skills --skill neural-audit

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 18 days oldThe repository was created 18 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

3.8 KB, 847 tokens by cl100k_base, as published. Nobody here has run it

Neural Portfolio Audit

First: Read the file ~/.claude/skills/seo-references/core.md before proceeding. Apply the core methodology and voice defined there to all output.

Also read: ~/.claude/skills/seo-references/data-pull-patterns.md for MCP tool parameter reference.

Step 1: Pull Portfolio-Wide Data

This skill has NO domain argument — it runs across the entire portfolio. Iterate through the MCP Routing Map in core.md and pull data for each domain that has live properties.

Semrush (PRIMARY — run for each domain):

  1. For each domain in the MCP Routing Map, call mcp__semrush__execute_report — Domain Overview (database=us). Captures: authority score, organic traffic estimate, keyword count. This is the most important data for the audit.

GA4 (ENRICHMENT — run for each property): 2. For each GA4 property in the MCP Routing Map, call mcp__google-analytics__run_report with metrics=[sessions, bounceRate], dateRanges=[{startDate: "28daysAgo", endDate: "today"}]. Captures: real traffic + engagement.

GSC (ENRICHMENT — run for each property): 3. For each GSC property, call mcp__gscServer__get_search_analytics with dimensions=[], rowLimit=1 (totals only). Captures: total clicks + impressions per property.

Note: This skill makes many MCP calls. If rate limits or timeouts occur, collect as much data as possible and note which domains had incomplete data.

Step 2: Analyze with Pareto Ruthlessness

Using the pulled data + the Full 13-Domain Roster (including LOCKED domains with CPC/volume data):

Section 1 — Total Addressable Market (TAM): Per domain: rough TAM = CPC × monthly volume × 12 × estimated CTR (use 3% for top-3 position, 1% for positions 4-10). Sum for total portfolio valuation.

Section 2 — Priority Matrix: Rank all 13 domains by: (CPC × Volume × Conversion Probability). Factor in:

  • Domain status: LIVE (can act now) > ACTIVE > ACQUIRED (needs setup) > LOCKED (blocked)
  • Vertical profit margins (legal = highest, home service = moderate, retail = lowest)
  • Current traffic (from GA4) vs. potential traffic (from Semrush) Identify the top 3 domains that represent 80% of revenue potential.

Section 3 — Attack Sequencing (THIS WEEK): For each of the top 3 domains, give 3 specific actions with exact deliverables:

  • Domain 1: [highest priority] — specific pages to create, specific keywords to target
  • Domain 2: [second priority] — specific actions
  • Domain 3: [third priority] — specific actions

Section 4 — Blind Spots & Risks:

  • Sleeping assets: LOCKED domains with high CPC × volume (what's being left on the table?)
  • Wasted effort: domains with low CPC in commoditized verticals
  • Portfolio concentration risk: what % of TAM is in the top 1-2 domains?

Section 5 — Purchase-Intent Page Deployment Status: Per domain: how many purchase-intent pages are deployed vs. how many are needed (based on topical map estimates of 13 pages per domain)?

Section 6 — No-BS Verdict: "Here's where you should put $1,000 in SEO budget this month. Here's why. Here's what you'll get."

Close with the one move that changes everything: the single highest-leverage action to take this week.

Step 3: Output

Follow the Output Protocol from core.md:

  1. Print the full audit playbook to terminal
  2. Extract structured summary
  3. Save to Graphiti with name Neural Audit — Portfolio and group_id {{GRAPHITI_GROUP_ID}}

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,851. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.