Skip to content
OpenAgentSkill
Browsing is open · Catalog captured 2026-11-21 · Automatic updates and account services are paused.
Marketing & sales

ip-purity-check

Check how "clean" an IP address is before using it — purity score (0-100), residential vs datacenter, native vs broadcast, proxy / VPN / Tor detection (with the VPN operator when known), public blocklists, abuse and attack history, and per-scenario suitability for AI services (ChatGPT, Claude, Gemini), social platform sign-up (X, Discord, Telegram), streaming, gaming platforms, cross-border e-commerce and email sending. Use when the user asks whether an IP or proxy is clean, safe to use, residential, a VPN or datacenter IP, why an IP keeps triggering captchas or risk control, which of several proxy IPs is best, or before an agent runs a network-sensitive task through its current egress IP. Powered by the free IPure API (ipure.dev), no API key required.

minner-fun · AI review on file

Installation

Read the source and dependencies before running a command. Copying does not install or execute anything.

npx skills add minner-fun/ipure-skills --skill ip-purity-check

About this skill


name: ip-purity-check description: Check how "clean" an IP address is before using it — purity score (0-100), residential vs datacenter, native vs broadcast, proxy / VPN / Tor detection (with the VPN operator when known), public blocklists, abuse and attack history, and per-scenario suitability for AI services (ChatGPT, Claude, Gemini), social platform sign-up (X, Discord, Telegram), streaming, gaming platforms, cross-border e-commerce and email sending. Use when the user asks whether an IP or proxy is clean, safe to use, residential, a VPN or datacenter IP, why an IP keeps triggering captchas or risk control, which of several proxy IPs is best, or before an agent runs a network-sensitive task through its current egress IP. Powered by the free IPure API (ipure.dev), no API key required. license: MIT

IP purity check (IPure)

IPure (https://ipure.dev) aggregates RDAP, BGP, cloud provider ranges, DNSBLs, the Tor exit list and commercial risk databases into one explainable report: every point deducted is traceable to a specific signal and data source. The API is free and needs no key.

When to use this skill

  • "Is this IP clean?" / "Is this proxy any good?" / "Is this a residential IP?"
  • "Can I use this IP for ChatGPT / Claude / X sign-up / Netflix / Steam / an Amazon store / sending email?"
  • "Why does this IP keep getting captchas / blocked / flagged?"
  • "Which of these proxy IPs should I use?" (compare several)
  • Before you (the agent) run a scraping, automation or account-related task: check the current egress IP first.

When NOT to use it

  • DNS configuration, website deployment, connectivity troubleshooting (ping / traceroute problems).
  • Questions that need a platform's internal risk data ("is my account flagged by Amazon?") — nobody outside the platform can see that.
  • Bulk scanning of IP ranges. The API is rate-limited and meant for individual lookups.

How to call

Look up a specific IP (IPv4 or IPv6):

curl -s "https://ipure.dev/api/lookup?ip=8.8.8.8"

Look up the caller's own egress IP — omit ip. This never requires verification, so it is the right call for "check my current IP":

curl -s "https://ipure.dev/api/lookup"

The full report is large. Pipe it through this filter to keep only what matters:

curl -s "https://ipure.dev/api/lookup?ip=8.8.8.8" | jq '{
  ip, reportUrl, queriedAt, stale,
  purity: .risk.purity, label: .risk.label, verdict: .risk.verdict, confidence: .risk.confidence,
  usageType, nativeType,
  vpnOperator: (.vpnOperator.name // null),
  country: .geo.country, asn: .asn.asn, org: .asn.org,
  flags: [.flags | to_entries[] | select(.value == true) | .key],
  factors: [.risk.factors[] | select(.points != 0) | {label, points, detail}],
  scenarios: [.scenarios[] | {id, score, levelLabel, reason}],
  blocklistsHit: [.blocklists[] | select(.listed and (.benign | not)) | .label],
  failedSources: [.sources[] | select((.ok | not) and (.skipped | not)) | .name],
  unknowns: [.unknowns[].label]
}'

If jq is unavailable, fetch the JSON and read the same fields. With Python use requests (or set a User-Agent); the default Python-urllib user agent is rejected by the CDN.

Reading the result

Text fields (label, verdict, detail, reason, levelLabel) are in Chinese — translate them into the user's language.

FieldMeaning
risk.purity0-100, higher is cleaner. 95+ pristine · 85+ clean · 70+ neutral · 50+ suspicious · 25+ risky · below 25 dangerous (risk.level)
risk.confidencelow / medium / high — data-source coverage. Separate from the score: the same 85 means less with two sources than with eight
risk.factors[]Where every point came from. points > 0 lowers purity, points < 0 raises it. floor marks decisive evidence (Tor exit, Spamhaus listing) that positive signals cannot offset
usageTyperesidential · mobile · business · hosting · education · government · unknown
nativeTypenative (registered and used in the same country) · broadcast (announced cross-region — platforms may see a location mismatch)
flagsisProxy, isVpn, isTor, isHosting, isRelay (iCloud Private Relay / WARP), … A flag confirmed by a single source is scored at half weight — see flagAgreement
vpnOperatorPresent only for known commercial VPNs (Mullvad, NordVPN, …): name, anonymity, logging policy, protocols
scenarios[]Suitability per use case, each with its own weighting: ai, social, streaming, gaming, ecommerce, email. Levels: excellent · good · fair · poor · unusable · restricted (the region itself is not served, e.g. AI services from mainland China)
blocklists[]listed = hit. benign = hit that is not a risk (Spamhaus PBL only declares a residential range). unavailable = the list could not be queried — unknown, not clean
abuse[]Abuse-report score and attacks.byType (login attempts, registration attempts, vulnerability probing…)
sources[]Per-source status. ok: false means that source failed — missing evidence, not a clean result
feedbackReal-world experience ratings from visitors who were actually using this IP, keyed by scenario id: count and average (1-5; 5 = works smoothly, 3 = usable with frequent captchas, 1 = does not work). A scenario appears only once it has 3+ ratings. Independent of the score — when present, quote it next to IPure's own verdict
unknowns[]What cannot be determined from an IP alone. Always pass these on
queriedAt, staleWhen the report was generated; stale: true means it is old enough that a re-check in the browser is worthwhile
reportUrlThe human-readable report. Cite it

How to present the answer

  1. Lead with the purity score, its level and the one-line verdict.
  2. Answer the user's actual question with the matching scenario (AI, social sign-up, streaming, gaming, e-c
Read the complete source ↗