A continuously updated feed of currently live phishing hosts, found and checked by an independent detection pipeline that I run on my own. I publish it as an ingest source for anti-phishing blocklists, DNS resolvers and wallet protection services. It is not affiliated with any of them.
This is a current-state list, not an archive. Each run republishes every host that is live right now. A host drops off automatically when its page goes dead or I withdraw the detection, so a consumer that replaces its copy on each pull never has to process removals.
| Slice | Plain hosts | With metadata | For |
|---|---|---|---|
| all | phishfeed_all.hosts |
phishfeed_all.json |
URL and host blocklists |
| crypto | phishfeed_crypto.hosts |
phishfeed_crypto.json |
wallet and crypto-scoped blocklists |
| dns | phishfeed_dns.hosts |
phishfeed_dns.json |
DNS resolvers that block a whole name |
Ingest URLs. Use these; they stay put whatever happens to this repository:
https://feed.risingsunsolutions.com/phishfeed_all.hosts
https://feed.risingsunsolutions.com/phishfeed_crypto.hosts
https://feed.risingsunsolutions.com/phishfeed_dns.hosts
The same files at this repository's raw address, updated by the same run:
https://raw.githubusercontent.com/jaym2026/phish-feed/main/phishfeed_all.hosts
https://raw.githubusercontent.com/jaym2026/phish-feed/main/phishfeed_crypto.hosts
https://raw.githubusercontent.com/jaym2026/phish-feed/main/phishfeed_dns.hosts
The .hosts files carry one hostname per line. Each .json file carries
generated_at (UTC) and count at the top, then one object per host in
hosts.
all: the host serves a live phishing page, confirmed either by my own check of the page or by at least two independent outside sources. Before anything is listed, these are removed:
- tutorial, portfolio and coursework clones
- sub-resources (images, scripts) that are not pages
- any shared-platform hostname where the abuse is only in the path, such as a shortener, a paste site or an object-storage endpoint. Listing those would block the whole provider.
crypto: the subset whose brand or hostname matches a wallet, exchange or DeFi term.
dns: the subset that is safe to block at the resolver. A URL blocklist
stops one page, but a resolver stops the whole name for every user behind it.
So this slice holds only names that exist to serve the phishing page. Each
entry's dns_reason says which of these three it is:
dns_reason |
Meaning |
|---|---|
platform_tenant |
a hostname that is one tenant's deployment on a shared platform (for example name.pages.dev), so blocking it affects nobody else |
young_domain |
a domain registered within the last 180 days |
root_kit |
the phishing page is the site itself, served at the root of a domain not known to be old |
A compromised legitimate site with a kit dropped in a folder is never in this slice, because blocking the name would take the victim's own site offline. It stays in all for URL-level blocklists.
The dns slice also applies a stricter evidence bar. Every entry needs all of the following:
- my own check of the page (outside listings alone are not enough)
- a named impersonated brand, or a listing by an independent phishing feed
- independent evidence beyond a sign-in form: the same bar I require before filing an abuse report, such as corroboration from outside sources or a brand read from the page itself rather than from the hostname
- no detection on the host that I have ruled a false positive
- a liveness check within the last 48 hours
- a page that is not a stock hosting control panel, a suspension notice or a server default page
| Field | Meaning |
|---|---|
host |
the hostname to block |
category |
phishing, malware or command_and_control |
brand |
the impersonated brand, when known (brands lists all of them when there are several) |
crypto |
true when the entry is in the crypto slice |
dns_reason |
why a resolver may block the name (see above) |
evidence_url |
the most recent URL seen serving the page on this host |
hosting_platform |
the hosting provider, when recognized |
verified |
true when my own check confirmed the phishing page |
corroboration_count |
how many independent outside sources also flag it |
source_feed |
the channel through which the host was first discovered |
first_seen / last_seen |
UTC, YYYY-MM-DD HH:MM:SS |
last_checked |
UTC time of the most recent liveness check |
If a host here should not be, please open an issue. This link opens one with
the title ready to edit (replace example.com with the hostname):
https://github.com/jaym2026/phish-feed/issues/new?title=False%20positive:%20example.com
I review reports and remove a confirmed false positive. Because the feed is current-state, the host is gone from the next hourly publish, so consumers do not need a separate retraction.
Republished automatically every hour. Every JSON file carries generated_at.
If the feed stops updating, a watchdog alerts me after 48 hours.
Open an issue: https://github.com/jaym2026/phish-feed/issues
Data is published for defensive use. No warranty: verify before enforcement.