More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Semantic Meaning & Intent-Gnostic Content Architect

    1

    100% offline semantic meaning & cognitive intent architect for Markdown/MDX, HTML5, and Gutenberg. Aligns DOM containers to 5 intents, audits H1-H6 propositions & RAG chunks, and synthesizes section-grounded img[alt] text.

    Free

    Security scanned
    Semantic Meaning & Intent-Gnostic Content Architect

    Semantic Meaning & Intent-Gnostic Content Architect

    Example session with this skill installed

    Audit and normalize our engineering post at https://kinetixcloud.io/engineering/pgbouncer-vs-supavisor-connection-pooling (post.mdx) where the draft skips from H2 to H4, uses a tripartite 'Pooling: PgBouncer, Supavisor & Odyssey' heading, opens with 'Direct Answer:', lacks a comparison table, and leaves raw 'IMG_4092.png' and stock-photo alt text on architecture diagrams.

    • Read your context and instructions
    • Compiled the semantic meaning

    1. Semantic Meaning & Cognitive Intent Scorecard

    • Detected Stack Format: markdown | Resolved Cognitive Intent: EVALUATIVE
    • Canonical Thesis (H1): PgBouncer vs Supavisor Connection Pooling Under Burst Traffic
    • Semantic Trajectory: 41/100 (TEMPLATE_SLOP_OR_INTENT_MISMATCH) -> 96/100 (SEMANTICALLY_ALIGNED)

    Pillar Breakdown: Intent Container 10/25 | Headings & Passages 13/30 | Visual & Link Semantics 10/30 | Schema & Crawl 8/15

    2. Diagnostic Findings & Structural Remediations

    PillarCodeSeverityFindingDeterministic Remediation
    intent_containerMISSING_EVALUATIVE_TABLEHIGHEvaluative intent lacks a structured comparison <table> or Markdown matrix.Add a comparison table with <thead> and a decision-oriented <caption>.
    heading_passageHEADING_HIERARCHY_SKIPMEDIUMSkipped heading levels fracture DOM outline: H2->H4 ('Memory Footprint per 10k Idle').Re-balance subheadings sequentially (H1 -> H2 -> H3).
    heading_passageFORMULAIC_HEADING_MONOTONYMEDIUMTemplate heading patterns (1): Tripartite H2: 'Pooling: PgBouncer, Supavisor & Odyssey'.Write single-concept propositions around one mechanism or trade-off.
    heading_passageLLM_SCAFFOLDING_OR_PURPLE_SLOPHIGHDetected LLM scaffolding/cliches (direct answer:).Strip 'Direct Answer:'/'Key Takeaways:'/purple prose; lead with concrete friction.
    visual_link_semanticsMISSING_OR_GENERIC_ALT_TEXTHIGHImage alt gaps: 0 missing, 1 generic/filename ('IMG_4092.png').Apply section-grounded 40-125 char alt text connecting subject to parent H2/H3 and H1.

    3. Section-Grounded Multimodal img[alt] Candidates

    Image SourceDefectParent SectionSynthesized Contextual Alt
    IMG_4092.pngGENERIC_OR_PREFIX_ALTTransaction Mode Tail LatencyTransaction Mode Tail Latency (PgBouncer vs Supavisor Connection Pooling Under Burst Traffic)
    latency-p99-chart.webpSEMANTIC_ALT_DRIFTMemory Footprint per 10k IdleLatency p99 chart illustrating Memory Footprint per 10k Idle

    4. Deterministic AST Normalization Log

    • Stripped bold snippet-bait opener prefix.
    • Re-balanced Markdown H4 -> H3: 'Memory Footprint per 10k Idle'
    • Synthesized Markdown alt 'Transaction Mode Tail Latency (PgBouncer vs Supavisor Connection Pooling Under Burst Traffic)'

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Stop Forcing One Rigid AI Template Onto Every Search Intent

    Most AI-generated drafts and CMS templates lose organic and Answer Engine Optimization (AEO) visibility before schema or backlinks ever matter: they stamp the same formulaic layout—a bold "Direct Answer:" blurb, a forced comparison table, tripartite A, B & C headings, skipped H2 -> H4 outlines, and stock-photo or filename img[alt] attributes—onto every URL.

    Search indexers and LLM RAG chunkers build semantic meaning from DOM structure and multimodal coherence: whether your HTML/Markdown containers match how a human resolves the query, whether subheadings form single-concept propositions supporting the H1, and whether every img[alt] bridges the visual asset to its parent H2–H4 section.

    Semantic Meaning & Intent-Gnostic Content Architect equips your coding agent with a 100% offline, pure-stdlib Python diagnostic and AST normalization suite (0 API keys, 0 pip packages, < 300 ms per file) built for Next.js / Astro Markdown & MDX, Semantic HTML5 / Shopify templates, and WordPress Gutenberg (<!-- wp:... -->) blocks.

    4 Deterministic Pillars Audited & Normalized

    1. 5-Archetype Cognitive Intent -> DOM Container Fit (25 pts): Resolves EVALUATIVE (<table> + decision <caption>), OPERATIONAL (<dl> Quick-Reference specs or <ol>), STRUCTURAL (architectural planes & failure modes), INSPIRATIONAL (bans forced comparison tables and <details> FAQ bloat in favor of sensory vignettes), and NAVIGATIONAL (verified primary-authority gateway anchors).
    2. Propositional Heading Tree & RAG Passage Saliency (30 pts): Enforces single H1 + sequential H1 -> H2 -> H3 descent, audits H1-to-subheading lexical support, flags tripartite noun-list headings, AI crutch prefixes ("Deep Dive into..."), explicit year-stamping, <h5>/<h6> photo-credit abuse, 4+ sentence RAG walls, robotic snippet-bait openers, and low empirical anchor density.
    3. Multimodal img[alt]-to-Page & Link Semantics (30 pts): Evaluates img[alt] coherence against H1 + parent H2–H4 + <figcaption>, synthesizes 40–125 char section-grounded alt candidates (scrubbing camera EXIF strings, UUID hashes, and IMG_0123 filenames), silences <= 48px decorative icons (alt="" aria-hidden="true"), audits inline <svg> accessibility, repairs blind image links, and synchronizes WordPress Gutenberg block JSON attributes with HTML tags.
    4. Read-Only Schema & Crawl Readiness Triage (15 pts): Audits <head> metadata, intent-eligible Schema.org @type coverage, multi-script JSON-LD split-brains, Wikidata QID entity grounding, and URL subdirectory depth.

    Requirements & Specialist Ecosystem

    • Requires Python 3.9+ (100% standard library; 0 pip dependencies, 0 external network requests).
    • Designed for inside-the-post semantic architecture and DOM normalization. When cross-system infrastructure issues surface, Pillar 4 cleanly triages when to pair with Schema Graph Harmonizer & De-Duplicator, Zero-Click Topic Vector & Intent Auditor, Crawl Budget & Indexation Diagnostician, or AEO Entity & Knowledge Graph Architect.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for E-commerce.

    Operates 100% offline on local Markdown, MDX, HTML, or Gutenberg files without crawling remote URLs, querying live Google SERPs, or stitching multi-plugin JSON-LD scripts (pair with our specialist paid skills for live SERP/Wikidata/JSON-LD/GSC workflows).

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 7 days ago

    • Free with an Agensi account

    Software
    Python 3

    Permissions

    Terminal / Shell
    Read Files
    Write Files

    Runs 100% offline using only the Python 3.9+ standard library. Zero external network requests, zero browser automation, and zero API keys required.

    Listed7 days ago

    What's inside

    Frequently Asked Questions

    It is tech-stack-agnostic at its core. It auto-detects Markdown/MDX (.md, .mdx with YAML frontmatter and ESM imports), full-page or fragment HTML5 (isolating main editorial content from nav/footer/sidebar chrome), and WordPress Gutenberg blocks (keeping <!-- wp:heading --> and <!-- wp:image --> JSON attributes synced with HTML tags).

    Search and multimodal RAG pipelines weight an image's surrounding DOM context heavily. Our deterministic engine combines cleaned semantic filename tokens (stripping UUIDs, timestamps, and camera EXIF strings), <figcaption> text, the active H2-H4 section proposition, and the canonical H1 thesis into a 40-125 character alt candidate—while silencing <=48px decorative icons with alt='' and aria-hidden='true'.

    Intent-gnostic content architecture rejects one-size-fits-all templates. While Evaluative queries require comparison tables and Operational queries require <dl> or <ol> specs, forcing rigid matrices and generic FAQ accordions onto Inspirational or cultural-discovery narratives triggers template-slop demotion and breaks reader immersion.

    Zero. It is 100% free, 100% offline, and uses only the Python 3.9+ standard library.

    New in E-commerce

    Free