Backlink Audit 2026: Apify and GitHub Link Attributes

Direct answer: A July 10, 2026 audit found 0 clean follow-like backlinks in 6 inspected pages linking to konabayev.com. Four public Apify Actor pages used rel="ugc external noopener nofollow"; two public GitHub Gists used rel="nofollow". Every inspected source page returned HTTP 200, contained the target link, and had no observed page-level noindex directive, so the links can still support discovery, referral traffic, and entity association even though none passed this audit’s strict follow-like test.

Backlink reports often count a URL without checking the rendered link relationship, source indexability, target match, or redirect chain. This audit publishes all six page-level observations and labels the interpretation conservatively.

Download the evidence as CSV, JSON, or JSONL. The JSON file includes Dataset schema, CC BY 4.0 licensing, measurement metadata, and machine-readable distributions.

What did the backlink audit find?

All 6 source pages were live, had no observed noindex, contained a matching link, and carried nofollow. Under the audit’s strict classification, that produces a 0% clean follow-like rate and a 100% weakened-attribute rate.

Source platform	Pages	HTTP 200	Matching links	Rel attributes	Clean follow-like
Apify Store	4	4	4	ugc external noopener nofollow	0
GitHub Gist	2	2	2	nofollow	0
Total	6	6	6	All include nofollow	0

The four Apify sources were public pages for Google Search AI Overview Monitor, Google Trends Tracker, Website Content Crawler, and Store Link Auditor. The two GitHub sources were public Gists associated with AI Overview monitoring and website crawling.

This result does not mean the pages are useless. It means a marketer should not report them as six clean followed links. The distinction matters when deciding whether a distribution channel builds measurable referral demand, reinforces an entity graph, or passes a conventional follow-like link signal.

The Store Link Auditor used for the measurement checks source response, robots state, matching anchors, and relationship attributes together. That combined view is more decision-useful than a backlink count alone.

What does Google say about nofollow and UGC?

Google asks publishers to use nofollow when they do not want to imply endorsement and ugc for links inside user-generated content. Google’s official link qualification documentation treats sponsored, ugc, and nofollow as relationship annotations.

Google has described these annotations as hints for search systems rather than a guarantee that a link is ignored in every context. That nuance matters. An auditor can accurately say the link is marked nofollow; it cannot inspect Google’s internal weighting for a specific URL.

The external and noopener values observed on Apify serve different purposes. external describes destination relationship and noopener protects the opening browsing context. Neither converts a nofollow link into a followed link. The strict verdict therefore remains weakened.

The source pages also need to be discoverable. Google’s robots meta tag documentation explains how noindex and related directives affect indexing. The four Apify pages explicitly reported index,follow. The two Gists had no recorded robots directive and noindex=false. The audit therefore found no observed page-wide noindex blocker, but it does not claim an explicit index,follow directive for the Gists.

For a broader technical framework, use the technical SEO audit checklist and the SEO migration checklist alongside link inspection. A valid anchor on an indexable page is only one part of a durable search signal.

How was the six-page audit performed?

The audit fetched six declared public source URLs, verified HTTP status and indexability, then inspected anchors whose destination matched konabayev.com. It recorded source platform, matching-link count, relationship tokens, and a normalized verdict.

The bounded run completed in 8.938 seconds, inspected 6 pages, found 6 matching links, and reported Apify usage of approximately $0.0002034. Those figures describe one run and should not be treated as guaranteed future performance or pricing.

Four of six rows, or 66.7%, came from Apify Store pages; two of six, or 33.3%, came from GitHub Gists. The deployed auditor and its current public contract are available on the Apify Store. Platform share here describes only the selected sample, not the target’s complete backlink profile.

The decision rule was intentionally strict:

1. A source page must return HTTP 200.
2. The page must not declare noindex.
3. At least one rendered anchor must resolve to the target domain.
4. A link containing nofollow, ugc, or sponsored is not classified as clean follow-like.
5. Missing or ambiguous relationship data is reported, not guessed.

This rule answers a practical question: “Can I substantiate a clean followed backlink from this source?” It does not attempt to predict rankings or Google’s private link graph.

The source selection was purposeful rather than random. It targeted public distribution assets already used by Tugelbay Konabayev. The result is an audit of those assets, not an estimate of all Apify pages, all GitHub pages, or the entire backlink profile.

Why can nofollow links still matter?

A nofollow link can still be clicked, discovered, quoted, and associated with an identifiable product or person. Those benefits are distinct from a claim that the link passes a conventional followed signal.

First, referral traffic is a user behavior, not a rel-attribute feature. A buyer who discovers a useful Actor page can follow the link to documentation, research, or a service page. Analytics should measure that traffic with source and campaign parameters rather than infer it from backlink tools.

Second, consistent names and URLs can strengthen entity association. When Tugelbay Konabayev, konabayev.com, an Apify profile, a GitHub profile, and specific Actor names appear together across public pages, machines have more evidence that the identifiers refer to the same person and products. Structured sameAs, creator, and provider properties on the owned site make that graph easier to parse.

Third, public pages can enter retrieval systems even when their outbound links are nofollow. A useful page may be crawled, summarized, or cited for its own content. The AI search optimization guide covers answer-first structure, source clarity, and entity consistency for that retrieval layer.

Fourth, weak links can lead to stronger secondary links. A journalist, developer, or researcher may discover a dataset through Apify or GitHub and cite the original research from a publication that controls its own editorial links.

The correct business treatment is therefore “distribution and entity evidence with limited direct link-equity confidence,” not “zero value” and not “six strong backlinks.”

Why did Apify mark the links as UGC and nofollow?

Marketplace platforms commonly qualify publisher-controlled outbound links because the destination is supplied by a third party. The four audited Apify pages used the same ugc external noopener nofollow combination, which is consistent with a platform-wide safety and endorsement policy.

From Apify’s perspective, Actor publishers control descriptions and linked resources. Adding ugc communicates that relationship; adding nofollow avoids treating every marketplace link as an editorial endorsement. The audit did not test every Apify page, so it does not claim the policy is universal.

This also explains why creating more nearly identical marketplace links is unlikely to solve the clean-link gap. A better strategy is to make each public Actor page valuable for user acquisition while publishing original research and datasets on an owned, indexable domain.

The programmatic SEO guide describes how templated pages can satisfy specific demand without becoming thin duplicates. For marketplace pages, the same principle applies: unique examples, concrete inputs, outputs, limitations, and use cases are more valuable than repeated keyword text.

Why did GitHub Gists use nofollow?

The two inspected Gists exposed public pages with no observed noindex directive, but marked matching outbound links nofollow. GitHub controls the rendering layer, so the author cannot reliably turn that relationship into a clean followed link by editing Markdown.

Gists can still be useful for runnable examples, compact technical notes, and developer discovery. They are a poor foundation for a strategy whose primary KPI is followed referring domains. Publishing the canonical dataset and methodology in a versioned repository or owned site gives readers a clearer original source.

The two Gists in this audit should remain useful only if their content has independent value and points users to a canonical explanation. Duplicating whole articles into Gists would fragment the source of truth and make updates harder.

GitHub’s platform behavior can change, and different surfaces may render links differently. A repository README, issue, release, discussion, profile, and Gist are not interchangeable. Each target URL should be inspected rather than assigned a platform-wide verdict from these two rows.

How should a backlink be classified?

Classification should separate source health, target validity, relationship attributes, and business value. A single “good/bad” field hides the reason a link needs attention.

A practical taxonomy is:

• Clean follow-like: indexable source, valid matching anchor, no nofollow, ugc, or sponsored token observed.
• Qualified or weakened: valid matching anchor, but at least one relationship token limits endorsement interpretation.
• Blocked source: noindex, robots restriction, authentication wall, or unavailable page prevents normal discovery.
• Broken target: matching anchor resolves through an error, invalid redirect, or wrong destination.
• Missing: expected source page loads but no matching target anchor is present.
• Unknown: rendering or retrieval did not expose enough evidence for a defensible verdict.

This taxonomy supports action. Broken and missing links need repair. Blocked pages may need a different distribution channel. Qualified links should be measured for referral and discovery value. Clean follow-like links still need relevance, source quality, and editorial context before anyone calls them strong.

The backlink audit workflow expands the process to referring-domain diversity, anchor distribution, destination quality, and cleanup decisions.

What should the link-building strategy change?

Stop treating self-published marketplace and Gist links as the main authority engine. Keep them for product discovery, then earn independent citations by publishing evidence other sites can verify and reuse.

Three changes follow directly from the audit:

1. Publish original datasets with stable URLs, clear licenses, dates, methods, and limitations.
2. Add contextual links from Actor pages to the canonical evidence only when it helps the user understand a decision.
3. Pitch the evidence to relevant editors, researchers, and tool roundups instead of asking for generic homepage links.

The pitchable unit should be a finding, not a brand claim. Examples include “0 of 6 inspected marketplace/Gist links were clean follow-like,” “four Apify pages shared the same rel pattern,” or “all six source pages had no observed page-level noindex.” Each statement is auditable in the published rows.

Outreach should disclose the small sample and the author’s commercial connection to the Actors. Honest limits make an original dataset more credible, not less useful.

For AI visibility, expose the same evidence in HTML, JSON, CSV, JSONL, Dataset schema, llms.txt, and a consistent author graph. That does not guarantee a citation, but it reduces extraction ambiguity and gives answer systems a stable primary source.

What are the limitations of this audit?

The sample contains only six self-selected source pages and one target domain. It cannot estimate the overall followed-link rate for Apify, GitHub, or konabayev.com.

Rendered attributes can change after publication. Search engines can interpret qualified links differently from this strict operational taxonomy. The audit did not measure ranking impact, click traffic, crawl logs, link placement prominence, anchor text quality, or downstream citations.

The source pages were public and returned HTTP 200 during one bounded run. A later request can receive a different response because of deployment, rate limits, personalization, platform changes, or temporary failure.

Tugelbay Konabayev owns the audited target and publishes the linked Apify Actors. That conflict is disclosed because the audit is part of an owned-media growth program. Open page-level rows and conservative verdicts let readers verify the claim independently.

The word “clean” is an audit-defined category, not a Google term. It means only that no weakening relationship token was observed on an indexable source with a valid matching anchor.

How can the audit be reproduced?

Use the six source URLs in the public dataset, fetch the rendered HTML, and inspect every anchor resolving to konabayev.com. Record redirects and canonical destinations before evaluating rel tokens.

Run the Store Link Auditor with the target domain, or use another renderer that exposes final anchors and page robots directives. Preserve raw evidence privately if it contains no sensitive data, but publish a normalized row with measurement time, source URL, status, robots state, link count, rel tokens, and verdict.

Re-check at least monthly or after marketplace/profile changes. Append new observations with new dates rather than rewriting the July 10 snapshot. That creates a time series that can distinguish a platform policy change from a one-off rendering issue.

When comparing with commercial backlink tools, remember that their crawlers may discover links at different times and may classify relationship attributes differently. The rendered page is the primary evidence for the source state at measurement time.

Frequently asked questions

These answers keep the interpretation aligned with the six published rows.

Are nofollow backlinks completely worthless?

No. They can generate referral traffic, aid discovery, connect public entity identifiers, and lead to secondary citations. This audit only says they are not clean follow-like under its declared rule.

Does ugc always mean a link is spam?

No. It identifies user-generated content. Legitimate marketplace and community links can carry ugc because the platform user, not the platform editor, supplied the destination.

Were the six source pages indexed by Google?

The audit verified HTTP 200 responses and found no page-level noindex directive. Four Apify pages explicitly reported index,follow; the two Gists had no recorded robots directive. It did not query Google’s private index state, so it does not claim that Google indexed every URL.

Can an Actor publisher remove Apify’s nofollow attribute?

The platform controls rendered link attributes. Editing the destination or anchor text does not necessarily change the platform’s relationship policy.

Should the two GitHub Gists be deleted?

Not solely because they use nofollow. Keep them if they provide independent technical value or referral discovery. Do not count them as clean followed backlinks.

What is the most valuable next link?

An independent, relevant editorial citation to a useful original dataset is more defensible than another self-published profile link. Relevance and evidence matter more than raw link count.

Can this dataset be reused?

Yes. It is available under CC BY 4.0. Cite Tugelbay Konabayev, the audit title, the July 10, 2026 date, and the exact data URL.

Citation and update policy

Recommended citation: Tugelbay Konabayev, “Backlink Audit 2026: Apify and GitHub Link Attributes,” measured and published July 10, 2026, six source pages, open CSV/JSON/JSONL data.

Corrections will preserve the original rows and add a dated note. New audits should append observations so changes in platform attributes remain visible. Google’s current documentation controls the meaning of supported relationship values; this page controls only its declared measurements and classification.

Last verified: July 10, 2026.