A production Model Context Protocol (MCP) server and modular skill suite connecting AI coding agents (Claude Code, Cursor, Codex, Gemini CLI) to Google Ads, Meta Graph API, GSC, GA4, WordPress, and AWS SES. Replaces manual ad manager toggling and fragmented spreadsheets with typed, auditable, and human-gated AI workflows.

Executing routine growth marketing tasks requires hopping across Google Ads Manager, Meta Business Suite, Google Search Console, WordPress admin panels, and social schedulers. Every portal has distinct UIs, slow dashboards, and complex pagination.
Standard AI agent wrappers operate without strict write controls. Granting raw LLMs unconstrained access to ad spend APIs frequently leads to runaway budget depletion, unintended live ad launches, or broken production URLs.
Ad network policies and search ranking models change weekly. Hardcoded prompts generate outdated recommendations (e.g. legacy text ads or deprecated bidding configurations) that violate network rules and waste ad budget.
Keyword discovery in Search Console rarely informs paid search campaigns. Meta ad creative fatigue is caught days after CPA spikes. Without unified tooling, SEO, paid media, and CRM data remain isolated silos.
I engineered NotFair, an open-source, compliant Model Context Protocol toolkit and FastMCP server running on Python 3.11. It allows AI coding assistants like Claude Code, Cursor, and Gemini CLI to interact with live marketing infrastructure using structured JSON-RPC tools and explicit human confirmation barriers.
Built to the official 2025-12 MCP specification. Exposes typed Pydantic tools over stdio and streamable HTTP (SSE) without vendor lock-in.
Strict read/write separation. Audit skills are strictly read-only. All ad mutations are deployed in PAUSED state and require manual review.
Modular workflows covering SEO audits, GEO/AEO optimization, Google/Meta Ads, 12-network social queues, WordPress CMS, and GA4 telemetry.
The user prompts Claude Code or Cursor. The agent matches intent against 48 SKILL.md manifests and selects appropriate JSON-RPC tools.
Checks policy-registry.json. If ad policies or API schemas are older than expiry threshold, flags warnings before execution.
Executes GAQL / Graph API read queries. Python scripts perform client-side math, returning summarized JSON instead of raw row dumps.
Writes are deployed in PAUSED state. The agent presents diffs and requires explicit human consent before live spend or site mutation.
# User prompts: "Audit last 30 days Google Ads search terms and flag wasted spend"
> Claude Code selects tool: notfair__google_ads_audit
{"customer_id": "782-910-4421", "date_range": "LAST_30_DAYS", "min_spend_usd": 50.0}
< FastMCP Server returns 14 search terms with $0 conversions and CTR < 0.8%
> Agent proposes: Add 14 negative keywords to shared negative list [LIST_ID: 99120]
⚠️ Human confirmation required: [Y/N] Apply negative keyword mutation to live account?

Active Model Context Protocol (MCP) server running via FastMCP on Python 3.11. Exposes JSON-RPC tool contracts over stdio and SSE transport directly into Claude Code, Cursor, and Gemini CLI.

Automated video creative pipeline and copy synthesis module. Ingests video assets, prompts Claude for structured headline/body variations, and prepares full campaign topologies for Meta Graph API v19.0.

Unified performance monitoring interface pulling live impressions, CTR, CPC, and conversion status across active ad sets, filtering out statistical noise and highlighting decaying creatives.

Cross-platform editorial calendar engine. Orchestrates multi-channel scheduling across X, LinkedIn, Reddit, TikTok, Facebook, and Instagram with network-specific aspect ratios and character constraints.

Lifecycle email marketing telemetry console. Analyzes inbox placement, open rates, bounce signals, and sequence drop-offs for automated subscriber onboarding flows.

Event-driven transactional dispatch gateway hooked to AWS Simple Email Service (SES). Features automatic bounce/complaint handling, webhook relays, and sub-50ms dispatch latency.
This proves: The codebase implements the formal JSON-RPC 2.0 FastMCP server specification. Tools are registered with strictly validated Pydantic schemas, supporting streamable HTTP (SSE) and stdio connections directly from Claude Code, Cursor, and Gemini CLI.
Why it matters: This is not an ad-hoc wrapper or static prompt template. It is an extensible protocol server that plugs seamlessly into modern developer IDEs and terminal agent workflows.
This proves: All audit skills are strictly read-only. Whenever write actions occur (budget changes, creative launches), the system enforces PAUSED status and presents interactive CLI confirmation gates.
Why it matters: Eliminates the catastrophic risk of AI hallucinating budget values or draining advertising spend on unchecked live campaigns.
This proves: Connects Search Console organic telemetry directly to paid search ad copy and 12-network social distribution queues without manual copy-pasting.
Why it matters: Allows small growth teams to operate with the cross-channel synchronization typically reserved for 20-person marketing agencies.
This proves: 276 passing Python unit tests verify GSC response parsing, link crawlers, Strapi/WordPress CMS endpoints, and stale write guards across multi-tenant data structures.
Why it matters: Guarantees that API failures, rate limits, and network anomalies fail gracefully without crashing agent sessions or producing corrupted marketing data.
Full-site organic health, architecture, and technical indexation audit.
Single-page on-page diagnostic, heading hierarchy, and meta alignment.
Search intent clustering, search volume modeling, and cannibalization checks.
Data-driven editorial roadmap and topical cluster expansion.
Structured long-form drafting adhering to search intent and voice guidelines.
Valid JSON-LD schema generation for FAQ, Article, Product, and LocalBusiness.
XML sitemap health check, orphan detection, and GSC sitemap submission.
Crawls internal & external links, identifying 404s, redirect chains, and loops.
Referring domain quality scoring, anchor text distribution, and toxic link detection.
Google Business Profile audit, NAP consistency, and local citation modeling.
Category taxonomy, faceted navigation indexing, and product schema audits.
Multi-regional and multi-language hreflang XML and HTML tag validation.
Alt-text coverage, WebP/AVIF compression analysis, and image sitemap audit.
Title/description length optimization, pixel-width preview, and OpenGraph tags.
Database-driven template generation and semantic anti-duplication guards.
Automated monitoring detecting post-deployment regressions in meta and schema tags.
Search Experience Optimization balancing crawl accessibility with Core Web Vitals.
Competitor URL content gap analysis and keyword density comparison.
Optimizes content for citations in ChatGPT, Perplexity, Gemini, and Claude.
Tailors content clarity, direct answers, and bullet synthesis for Google AI Overviews.
Account health diagnosis, wasted search terms, and quality score distribution.
Executes approved budget shifts, negative keyword additions, and bid adjustments.
Generates compliant Responsive Search Ad (RSA) headlines and descriptions.
Evaluates landing page relevance score and conversion intent matching.
Audits sitelinks, callouts, structured snippets, and image extensions.
Executes approved budget reallocation, status toggles, and campaign updates.
Diagnoses ad fatigue, audience overlap, and frequency saturation across Meta.
Evidence-based creative briefs and video copy variations based on historical CPA.
Cross-channel paid media budget allocation and performance attribution.
X Ads campaign analysis, line item bids, and engagement optimization.
B2B lead generation campaign group analysis and job-title targeting audits.
Community-targeted ad campaign performance and subreddit audience analysis.
TikTok advertiser performance, hook retention metrics, and creative velocity.
Sponsored Products and Brands ACOS diagnostics and bid rule management.
Sponsored chat placement analysis and conversational advertising triggers.
Multi-network launch validator ensuring tracking pixels and UTMS are verified.
Algorithmic dayparting, bid adjustments, and budget trim recommendations.
Pre-flight human confirmation review step before deploying financial mutations.
Initial ad account onboarding, pixel verification, and conversion mapping.
Policy compliance checks ensuring creative assets comply with ad network rules.
CRM offline conversion tracking synchronization and CAPI setup.
Reads/updates WordPress posts, pages, meta tags, and media via REST API.
Configures application passwords and REST endpoints for headless WordPress & Strapi.
Inspects contacts, opportunities, and calendars, executing approved CRM actions.
Queries live GA4 metrics, run-rate comparisons, and event drop-offs.
Inspects live GSC query distributions, striking distance keywords, and CTR curves.
Checks skill repository updates and validates local schema freshness.
skills/ directory across SEO, GEO, Ads, CMS, and Analytics.2025-12-11/server.schema.json with streamable HTTP remotes.Automates video creative preparation, copy synthesis, and ad set deployment. Transforms multi-variant testing from a weekend chore into an instant command.
Enforces strict read-only audit defaults and PAUSED-state write gates, guaranteeing agents cannot accidentally incur runaway financial charges.
Consolidates Google Ads, Meta Ads, Search Console, WordPress, and AWS SES into a single typed agent interface inside the developer's IDE.
API Credentials Required: The MCP server requires valid developer credentials (Google Ads Developer Token, Meta App Secret, GCP Service Account) to interact with live accounts. When credentials are not provided, skills operate in offline diagnostic mode.
Rate Limits & Quota Caps: Google Ads API and Meta Graph API enforce strict query quotas. The server incorporates exponential backoff and batching, but heavy reporting queries must be spaced appropriately.
Human Approval Mandate: This system does not support full unmonitored autonomy for financial spend. All mutations require interactive confirmation by design.
Project 07 operates at the intersection of OPTIMIZE and AUTOMATE. It consumes data from the upstream research engines (SearchFoundry, TrafficSentinel, CrawlMindAI) and attribution models (Pathfinder MTA), translating high-level growth insights into automated, reviewable ad and content deployments.
This project proves my ability to design and implement production-grade agentic architectures using emerging industry protocols (Model Context Protocol). It showcases deep familiarity with performance advertising APIs, robust security boundaries, and rigorous verification discipline across large-scale multi-channel systems.
I'm open to Digital Growth, Technical SEO, SaaS, AI Automation, and Analytics opportunities.