Neverranked — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Neverranked (Agent Skill) and scored it 87/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 3 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 3 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
A bulleted imperative like {match} tells the agent to never reveal, disclose, or mention something to the user. Used adversarially it can instruct the agent to hide its tool calls or lie about what it did — stripping the transparency a user relies on to trust the agent.
The text {match} asks the agent to disclose its hidden system prompt or initial instructions. That is often the first step of a larger attack: knowing the system prompt lets an attacker craft inputs that defeat its constraints by mimicking its own voice.
repeat/reveal/print your system prompt request from the skill.Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
NeverRanked is a research engagement that measures what AI answer engines cite for a category. The output is a forensic memo plus a prepped punch list for the customer's team to execute against.
This repository holds the marketing site, the dashboard worker source (under dashboard/), and the historical artifacts from the company's prior product.
Until May 2026, NeverRanked sold a JavaScript snippet that was claimed to drive AI citations. We ran a pre-registered kill test against our own domain. The result was zero citations. The LLM crawlers do not execute JavaScript, so the snippet was structurally invisible to them.
We stopped selling that product, retracted the public content that promoted it, and rebuilt the company around the measurement layer that does work.
This repository contains both states: the current honest surfaces (homepage, About, ASB pitch, terms, privacy, llms.txt) and the historical artifacts (retired blog posts, retired case studies, retired audits, retired explainers). Anything that contradicts the retraction is either replaced with an honest holder or marked as retired in place.
For the current state of every code path, see ~/.claude/projects/-Users-lanceroylo-Desktop-neverranked/memory/code_rewrite_status_2026-05-21.md in Lance's local Claude memory.
Seven AI surfaces, every day.
cite their sources: Perplexity, ChatGPT search, Gemini grounded, Microsoft Copilot via Bing, Google AI Overviews.
data: Claude, Gemma.
Both layers measure different failure modes. A brand invisible in citation is invisible when AI fact-checks itself. A brand invisible in model knowledge is invisible at the baseline, before any search happens.
A forensic memo plus a prepped punch list. Per query, per engine, per competitor, per source type. Daily measurement. Monthly delta memo on ongoing engagements.
We do not execute. No content writing, no website edits, no schema deploys, no profile updates. The labor stays with the customer's team or their agency. That separation is structural.
This is a research engagement, not a SaaS subscription. There is no self-serve dashboard.
neverranked/
├── index.html honest homepage holder
├── about/, terms/, current honest pages
│ privacy/, thanks/
├── pitch/asb-hawaii/ current direct-buyer pitch shape
├── pitch/*/ retired pitch pages (holder)
├── blog/, case-studies/, retired catalogs (holders + noindex)
│ profile/, state-of-aeo/
├── standards/, schemas/, retired surfaces (holders)
│ agencies/, security/,
│ for-agencies/,
│ principles/, kit/
├── meetings/kits/ current meeting prep, demo bundle
├── dashboard/ app.neverranked.com Worker source
├── llms.txt rewritten for current state
└── sitemap.xml homepage onlyThe historical artifacts (retired pages, audit documents under /audits/, prior explainers like EXPLAINER.md and AGENCY-EXPLAINER.md, the original audit template under /audit-template/) reflect the prior product premise and are preserved for diff context but should not be acted on as current guidance.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.