More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    Site-Wide Internal Linking Mesh Architect

    1

    Screaming-Frog-independent internal & external link mesh engine. Computes DOM-isolated PageRank, BFS click depth, entity-guarded broken-slug repairs, Google Link Spam compliance, and copy-ready bidirectional unified diffs.

    $29

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Site-Wide Internal Linking Mesh Architect

    Site-Wide Internal Linking Mesh Architect

    Example session with this skill installed

    Run a full site-wide internal & external link mesh audit on https://kinetixcloud.io (or ./content MDX repo) centered on our under-linked target post /engineering/pgbouncer-connection-pooling/. Surface nav-masked orphans, broken internal slugs, commercial affiliate rel='sponsored' violations, and compile copy-ready bidirectional link diffs.

    • Read your context and instructions
    • Compiled the site-wide internal linking

    1. Site-Wide Graph & External Compliance Telemetry (kinetixcloud.io)

    • Composite Mesh Health Score: 74/100 (Rate-Normalized Topology 68/100 | External Hygiene 86/100)
    • Indexed Editorial Nodes: 184 documents (<nav>/<footer>/sidebar chrome stripped prior to edge scoring)

    Editorial Body Edges: 412 links (2.24 mean inlinks/page) |

    Nav-Masked Orphans: 11 (in_degree_nav > 0, in_degree_body == 0)

    • Crawl Depth >= 4 Hazards: 6 pages | Broken Internal Slugs: 3 (entity-guarded repair candidates ready)

    2. Broken Internal Links & Entity-Guarded Fuzzy Repairs

    Source DocumentBroken Internal HrefVerified Catalog TargetConfidence
    /docs/postgres-overview//engineering/pgbouncer-pooling//engineering/pgbouncer-connection-pooling/89%
    /blog/read-replicas-scaling//guides/failover-testing-guide//guides/automated-failover-testing/76%
    /guides/cloud-billing//pricing/team-tier/(Contradiction Guard: /pricing/free-tier/ rejected)HOLD

    3. External Outbound Authority & Google Link Spam Audit

    Domain Tiers (318 external links): 24 primary_docs_ref | 12 statutory_gov_edu | 261 editorial_citation | 21 commercial_affiliate

    [HIGH] MISSING_REL_SPONSORED on /guides/cloud-hardware-benchmarks/ (https://retail-partner.io/nvme-ssd?ref=kinetix): Add rel="noopener noreferrer nofollow sponsored".

    [HIGH] BROKEN_EXTERNAL_LINK (HTTP 404) on /docs/tls-certificates/ (https://legacy-ca.org/crl-spec): Replace with verified standards archive.

    4. Bidirectional Mesh Blueprint for /engineering/pgbouncer-connection-pooling/

    DirectionPaired PagePageRank / StateScoreHook Mode & Surgical Diff Anchor
    INBOUND/guides/postgresql-high-availability/PR 84.20.81inline_wrap -> wrap outside <a>: [PgBouncer connection pooling](/engineering/pgbouncer-connection-pooling/)
    INBOUND/docs/kubernetes-statefulsets/PR 71.50.69inline_wrap -> wrap sentence span: [transaction pooling](/engineering/pgbouncer-connection-pooling/)
    OUTBOUND/guides/rds-proxy-cold-starts/ORPHAN (0 inlinks)0.74inline_wrap -> orphan rescue anchor [RDS Proxy cold starts](/guides/rds-proxy-cold-starts/)

    5. Coalesced Unified Diff Ready (--apply-diff Opt-In)

    --- a/guides/postgresql-high-availability.mdx
    +++ b/guides/postgresql-high-availability.mdx
    @@ -42,7 +42,7 @@
    -Most multi-tenant clusters stabilize idle connections via PgBouncer connection pooling before scaling read replicas.
    +Most multi-tenant clusters stabilize idle connections via [PgBouncer connection pooling](/engineering/pgbouncer-connection-pooling/) before scaling read replicas.
    

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

    About this skill

    Desktop Crawlers Stop at CSV Spreadsheets. This Skill Compiles the Surgical Patch.

    Most internal-linking tools and prompt wrappers force you to crawl your own site in a desktop GUI, export massive Screaming Frog CSVs, and still spend hours manually opening tabs to figure out which paragraph can actually host a link. Worse yet, naive crawlers count <header>, <footer>, share bars, cookie banners, and recent-post widgets as links—masking critical Editorial Orphans (in_degree_nav > 0, in_degree_body == 0) that lose all authority once they rotate off the homepage.

    Site-Wide Internal Linking Mesh Architect replaces manual crawler exports with a 4-module, 100% pure-Python 3.9+ standard-library CLI engine (0 pip dependencies, 0 per-URL crawler license caps, 0 third-party SaaS lock-in). Feed it a local Astro/Next.js/Hugo Markdown repo, a public XML sitemap, or a WordPress REST endpoint, and it audits your entire domain topology while compiling copy-ready unified .diff patches for any single target page.

    7 Forensic Modules Engineered for Curated Editorial & Docs Scale

    1. DOM Chrome Isolation (body vs. nav Partitioning): Strips <header>, <nav>, <footer>, <aside>, ARIA navigation/banner/contentinfo/complementary, cookie banners, .sharedaddy/.a2a_kit sharebars, and related-post widgets before scoring link equity—immediately surfacing Nav-Masked Editorial Orphans.
    2. Rate-Normalized Internal PageRank ($d=0.85$) & Archive-Aware Homepage BFS: Runs 35-iteration damped PageRank over editorial body links, merges rendered homepage / links with Dijkstra paginated archive traversal (click_depth >= 4 alerts), and uses density-normalized scoring that avoids 0/100 saturation on 200+ page websites.
    3. Entity-Guarded Broken Internal Slug Repair: Detects broken internal <a href> targets and pairs Jaccard + sequence similarity (>= 0.68) with a distinctive-entity contradiction guard so shared URL suffixes never mis-map conflicting topics or geographies.
    4. True-Cased Multiplicative Bidirectional Mesh Solver: Given any single post/page (--target), applies corpus true-casing and multi-category spam protection (len(categories) >= 4) before gating non-linking site articles on verified BM25 + Heading/Entity + Narrow Taxonomy similarity, eliminating cross-topic orphan leakage.
    5. Outside-<a> inline_wrap Hook Finder & Anchor Shield: Extracts hyphen-safe multi-word entity spans verified to sit outside existing <a> tags for natural in-sentence hyperlinking (inline_wrap), blocks common-noun anchor traps, flags over-concentrated exact-match anchors (> 30%), headings containing links, and blind image anchors.
    6. Chrome, Security & Tier-Exempt External Compliance Firewall: Classifies outbound citations (statutory_gov_edu, primary_docs_ref, editorial_citation, commercial_affiliate), recognizes rel="ugc", exempts authoritative citations (.gov/.edu, official docs, and creativecommons.org/licenses/ + /publicdomain/ deeds) from equity caps, detects affiliate params missing rel="sponsored", sorts HIGH -> MEDIUM -> LOW before report slicing, flags EXTERNAL_EQUITY_DILUTION (--max-equity-ratio-floor) & EXCESSIVE_EXTERNAL_DENSITY (--max-density-per-1k), and probes live HTTP 404/410/redirect chains (--probe-external).
    7. Coalesced Multi-Edit Unified Diffs (--apply-diff Opt-In): Compiles broken-slug fixes, inbound donor hooks, and outbound orphan-rescue links into atomic unified .diff patches and WordPress Gutenberg (<!-- wp:paragraph -->) replacements with html.unescape() safety.

    Supported Scale & Page Count Limits

    • Recommended Sweet Spot: Up to 5,000 pages (--source local / --source wp-rest) or 80–500 pages (--source sitemap, which defaults to --limit 80).
    • Practical Upper Ceiling: Up to 10,000 pages on local .md/.mdx/.html repos and WordPress REST (1,500 categories/tags max), and 1,500–2,500 pages for sequential live --source sitemap crawls (25 unique external URL probes/run by default).
    • Enterprise / Large Sites (> 10,000 pages): Scope audits to a sub-directory (--root ./content/guides) or pillar slice (--limit 2000) rather than ingesting huge tail UGC archives in one pass.

    Part of the Full-Stack Technical SEO & AEO Mesh Suite

    Pairs cleanly with Semantic Meaning & Intent-Gnostic Content Architect (intra-post headings & RAG chunks, Free), Crawl Budget & Indexation Diagnostician (depth >= 4 taxonomy queue traps), Zero-Click Topic Vector & Intent Auditor (cannibalization splits), Schema Graph Harmonizer & De-Duplicator, and 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.

    Built for curated editorial sites, blogs, and SaaS docs hubs—not web-scale million-page UGC platforms. Recommended sweet spot: up to 5,000 pages via local repo or WP REST (10,000 pages practical upper ceiling; WP taxonomy fetch caps at 1,500 terms) and 80–500 pages via live XML sitemap (1,500–2,500 practical upper ceiling with polite 0.25s sequential GETs; defaults to --limit 80). Live --probe-external checks default to 25 unique URLs per run. For >10,000-page sites, scope audits to a sub-directory (--root) or pillar slice (--limit).

    Does not export private Google Search Console metrics or run headless JavaScript browsers; dynamic SPA client-side routes without pre-rendered HTML or sitemaps should be audited against built static output (--source local). External live status checks (--probe-external) depend on outbound network access and rate-limit polite delays.

    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

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 days ago

    Software
    Python 3

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files

    Allowed Hosts

    Target site domain (only for --source sitemap/wp-rest)
    External outbound URLs (only for --probe-external)

    All 4 Python 3.9+ scripts run 100% offline with zero network calls when auditing local Markdown/MDX/HTML directories (--source local). Network access is used exclusively for --source sitemap/wp-rest or optional --probe-external HTTP HEAD/GET verification. File edits occur only on explicit opt-in --apply-diff within --root.

    Listed6 days ago
    Updated5 days ago

    What's inside

    Frequently Asked Questions

    Desktop crawlers output flat CSV link tables and force you to manually hunt through every article for placement context—while mixing navbar/footer menu links with body links. This engine strips DOM chrome automatically, isolates Nav-Masked Editorial Orphans, applies multiplicative topical gating so high-PageRank hubs never cross-link unrelated geographies, and compiles coalesced paragraph-level unified diffs ready to apply.

    Two built-in safeguards validated in large multi-topic catalogs: (1) fuzzy slug repair enforces a distinctive-entity contradiction guard so shared URL suffixes never mis-map distinct cities or products, and (2) multiplicative topical gating scales PageRank and orphan boosts by verified BM25 + entity + taxonomy similarity rather than adding free points to unrelated orphans.

    Zero API keys and zero pip packages are required. You can run full audits on local .md/.mdx/.html directories 100% offline, scan public XML sitemaps without credentials, or query WordPress REST endpoints (/wp-json/wp/v2/)—supporting optional WP Application Password auth via environment variables for draft/edit context.

    It classifies external domains into 5 authority tiers, detects affiliate/referral parameters (ref=, aff=, tag=, irclickid=) and commercial redirect paths missing rel='sponsored' or rel='nofollow', flags target='_blank' links missing rel='noopener', spots http:// downgrades and naked URL anchors, and probes live HTTP 404/410/redirect chains.

    Recommended sweet spot is up to 5,000 pages on local Markdown/MDX/HTML repos or WordPress REST (10,000 pages practical upper ceiling; up to 1,500 WP categories/tags) and 80–500 pages for live XML sitemap crawls (1,500–2,500 pages practical upper ceiling; defaults to --limit 80). For >10,000-page sites, scope audits via --root sub-directories or --limit slices.

    New in E-commerce