sergekostenchuk/seo-llm-skill-cluster / skills/semantic-core-architect
semantic-core-architect
VerifiedBuild evidence-labeled semantic cores for SEO, LLM-readable architecture, and site information design. Use this skill when the user asks for semantic core, query clusters, search intents, audience segments, entity/topic maps, language or locale priorities, competitor/source evidence, pillar-topic planning, or a structured handoff before URL architecture and internal linking. It must mark unknown search volume, difficulty, and ranking data as unknown instead of inventing metrics.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 3.9 KB |
| agents/openai.yaml | config | 410 B |
| assets/entity-topic-map.template.yaml | config | 839 B |
| assets/semantic-core.template.yaml | config | 1.3 KB |
| evals.json | config | 2.1 KB |
| references/semantic-core-rubric.md | markdown | 3.3 KB |
| validation-report.md | markdown | 1.3 KB |
references/semantic-core-rubric.md
View on GitHub# Semantic Core Rubric
## What Good Looks Like
A semantic core is a strategic map of demand, intent, entities, topics, audiences, languages, and evidence. It is not a raw keyword dump.
Good artifacts:
- separate observed facts, user-provided facts, inferred ideas, and unknowns;
- group queries by user intent and page role;
- identify entities that deserve stable canonical pages;
- include multilingual priorities without assuming every language launches at once;
- expose evidence gaps;
- prepare a clean handoff to URL architecture and link graph work.
## Required Cluster Fields
Each cluster should include:
- stable `id`;
- `label`;
- `primary_intent`;
- `audience`;
- `priority`;
- `languages`;
- `queries`;
- `entities`;
- `topics`;
- `evidence`;
- `unknown_metrics`;
- `downstream_notes`.
## Intent Categories
| Intent | Meaning | Example |
| --- | --- | --- |
| informational | User wants explanation or news context. | "AI news for builders" |
| navigational | User wants a known brand, author, project, or community. | "mlllm tg-news" |
| commercial | User compares tools, vendors, platforms, or approaches. | "best AI agent monitoring tools" |
| transactional | User wants to buy, subscribe, download, install, or join. | "subscribe AI news telegram" |
| support | User wants instructions, troubleshooting, or policy details. | "how to read llms.txt" |
## Evidence Labels
Use these source labels:
- `observed`: fetched page, logs, Search Console export, keyword tool export, or source document was directly inspected.
- `user_provided`: supplied by the user and not independently verified.
- `inferred`: reasoned from available facts.
- `unknown`: not verified.
Do not convert `inferred` into numeric volume or difficulty.
## Priority Scoring
Assign priority using this qualitative model:
| Factor | Signal |
| --- | --- |
| identity fit | The cluster describes what the site should be known for. |
| audience fit | The cluster serves the primary reader or buyer. |
| architecture value | The cluster affects URL structure, pages, or link graph. |
| evidence strength | There is observed or user-provided evidence, not only a hunch. |
| feasibility | The site can realistically publish useful content for the cluster. |
Use:
- `P0`: core identity and architecture driver.
- `P1`: important supporting cluster.
- `P2`: expansion cluster.
- `P3`: backlog or speculative.
## Entity Rules
Entities should have:
- stable lowercase id;
- human-readable name;
- type;
- canonical candidate URL when known;
- relationship to the site;
- evidence.
Allowed common types: `Organization`, `Person`, `Project`, `Product`, `SoftwareApplication`, `Topic`, `Publication`, `Community`, `Vendor`, `Technology`.
## Language Rules
For multilingual sites:
- define the default language;
- list launch languages separately from future languages;
- do not create equal priority for every language by default;
- note translation risk and review needs;
- identify locale-specific query differences.
## Handoff To IA
The semantic core may suggest page candidates, but final URL/canonical decisions belong to `information-architecture-seo`.
Handoff should include:
- clusters that need index pages;
- entities that need canonical pages;
- topics that could become pillar pages;
- language groups;
- duplicate-content risks;
- unknown data needed before final architecture.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |