Discover
Find seeds from competitor gaps, audience pain points, long-tail questions, adjacent topics, and existing content exclusions.
SEO / AEO / GEO
AI marketing workflow
Public proof artifact
AI search visibility, run like an operating system.
Jet SEO turns keyword research, Ahrefs exports, SERP analysis, content briefs, MDX generation, internal links, and validation into one auditable agent workflow. The point is not autonomous content spam. The point is search visibility work with evidence, state, and review gates.
batch-then-advance
one Ahrefs human handoff
stateful agent runbook
What the repo does
The public repo is a workflow pack. It contains the runbook, memory state, step agents, and content generation skill that let an AI coding agent execute an SEO content pipeline without losing context halfway through.
Validate
Prepare Ahrefs sessions, consolidate CSV exports, normalize columns, and preserve source-of-truth rows.
Prioritize
Score each keyword by business value, traffic potential, difficulty, funnel stage, search intent, content type, pillar, and cluster.
Analyze SERP
Capture top URLs, title patterns, competitor strengths, content format, and opportunity type before a brief is generated.
Create Briefs
Turn approved opportunities into structured briefs with outline logic, SERP evidence, and internal link targets.
Generate MDX
Use the atlas-seo skill to write in one pass with SEO, AEO, GEO, voice, persuasion, and AI-writing checks.
Link
Add hub-and-spoke internal links and backlinks so new posts join the existing content system.
Validate
Block low-quality output with SEO checks, AEO/GEO structure checks, frontmatter sync, and AI-writing tell detection.
Marketing-centric, not content-centric
The stronger Stripe-style story is workflow transformation. Jet SEO is valuable because it turns marketing judgment into a repeatable operating loop with state, evidence, review, and handoff.
Every seed moves through structured validation instead of becoming a one-off blog prompt. The workflow tracks what exists, what is deferred, and what is approved.
The atlas-seo skill encodes writing style, SEO basics, AEO/GEO answer structure, and AI-writing avoidance into the generation step.
MEMORY_STATE.md makes the workflow resumable after compacted context, long runs, or handoff to a new agent.
Operating details
| Layer | What it proves | Risk it handles |
|---|---|---|
| Seed discovery | Agent can research the opportunity space before writing. | Prevents generic content calendars and duplicate topics. |
| Ahrefs handoff | Human-in-the-loop step for paid data and export quality. | Prevents fake keyword metrics and uncontrolled credit spend. |
| Prioritization | Marketing strategy enters the system through business value, intent, and clustering. | Prevents traffic-only prioritization. |
| SERP analysis | Briefs are grounded in ranked pages and observed formats. | Prevents unsupported claims about what Google rewards. |
| MDX generation | Output is formatted for a codebase, not pasted into a CMS manually. | Prevents handoff loss between content and engineering. |
| Validation | The workflow blocks AI tells, missing SEO fields, weak AEO/GEO structure, and broken sync. | Prevents confident publishing of low-trust content. |
Interview story
30-second pitch: I built Jet SEO as a proof-of-work repo for AI-native marketing operations. It is a stateful agent workflow that takes search ideas through Ahrefs validation, SERP analysis, briefs, MDX generation, internal linking, and quality checks. The key lesson is that marketing AI should not be framed as "make infinite content." It should be framed as a reliable operating system for search visibility work.
Truth boundary: This artifact proves workflow architecture and public implementation detail. It does not claim ranking gains, traffic growth, or revenue impact without analytics evidence.
Sources used for market grounding
These sources informed the market-positioning layer. The implementation details come from the Jet SEO repo itself.