This is a master key left under the doormat, but only for the few doors that are actually exposed
CVE-2026-28777 is a hardcoded/trivial credential issue in the International Datacasting Corporation (IDC) SFX2100 Satellite Receiver: an unauthenticated attacker who can reach SSH can log in with the built-in user account using the weak password 12345, then break out from the restricted shell into a normal interactive session. NVD maps the issue to datacast:sfx2100_firmware and datacast:sfx2100 CPEs, but no narrower affected firmware range or fixed release has been published, so defenders should assume the exposed SFX2100 fleet is affected until proven otherwise.
The vendor/NVD 9.8 is too hot for real patch triage because it scores the flaw in a vacuum. In practice, the decisive friction is *reachability*: this is only exploitable where SSH on a specialized satellite receiver is reachable from the attacker position, and most enterprise/OT deployments put that management plane on private or segmented networks; that keeps this out of CRITICAL for fleet-wide prioritization even though the credential itself is embarrassingly weak.
4 steps from start to impact.
Find a reachable SFX2100 SSH service
nmap or masscan is enough to locate TCP/22 or whatever translated management path exists, and the public manual explicitly discusses remote network access for the Web GUI and advanced management functions on the SFX2100 platform.- Attacker has network path to the receiver management interface
- SSH is enabled and reachable from the attacker's segment
- Many deployments place these receivers on private management VLANs or OT enclaves rather than the public internet
- ACLs, jump hosts, VPN requirements, or serial/OOB-only administration can remove remote reachability entirely
Authenticate with the trivial built-in password
OpenSSH, the attacker logs in with user:12345 as documented by the researcher and reflected in the CVE/NVD record. No exploit development is needed; this is straight credential abuse against a built-in account.- Built-in
useraccount still exists - Password was never changed or account never disabled/locked
- Organizations that rotate the account password during commissioning kill this CVE outright
- Some deployments may disable SSH or disable the account even if the firmware remains vulnerable
Escape the restricted shell
- Successful login as the built-in account
- Restricted shell escape paths remain available
- This CVE by itself does not grant root
- Impact is bounded to what the
useraccount can access unless chained with other local issues on the appliance
last/wtmp, process listings, or north-south SSH session monitoring.Operate from the appliance foothold
- Interactive shell on the appliance
- Useful credentials, configs, or adjacent trust paths exist on the device
- Niche appliance role limits blast radius compared with a general-purpose VPN gateway or identity tier asset
- Further escalation or lateral movement depends heavily on local environment and adjacent trust
The supporting signals.
| In-the-wild status | No confirmed active exploitation found in available sources as of 2026-06-01; OpenCVE flags exploitation as none and CISA KEV does not list the CVE. |
|---|---|
| Proof-of-concept availability | Public exploit details are effectively available in Abdul Mhanni's disclosure, which names the credential as user:12345 and describes the SSH foothold. This is not a sophisticated exploit; any SSH client is the weaponized tool. |
| EPSS | 0.00435 (~0.435%) from the user's intel block; that is low and consistent with a niche device plus uncertain exposure population. |
| KEV status | Not KEV-listed. A direct CISA site search for CVE-2026-28777 returned no result, and OpenCVE also marks KEV as No. |
| CVSS vector reality check | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H captures the *technical* ease, but it overstates *fleet urgency* because it assumes attacker reachability to SSH on a specialized receiver. |
| Affected versions | Public records do not narrow this to a specific firmware build. NVD lists datacast:sfx2100_firmware and datacast:sfx2100; treat all deployed SFX2100 units as suspect until validated locally. |
| Fixed version | No patched version publicly identified. The researcher states IDC did not respond during the disclosure window, and OpenCVE notes no vendor fix/workaround was provided. |
| Scanning / exposure data | I found no trustworthy public Shodan/Censys/GreyNoise population number for this CVE or platform in the sources reviewed. That absence itself is meaningful: this looks like a niche, low-population appliance rather than a mass-exposed perimeter platform. |
| Disclosure date | 2026-03-04 in NVD/OpenCVE; the researcher blog was published 2026-03-05. |
| Researcher / reporting org | Finder credited as Abdul Mhanni; assigner/source is Gridware. |
noisgate verdict.
The single biggest downward pressure is attacker position: this only matters where someone can actually reach SSH on a specialized receiver, which usually means an internet-exposed management plane or prior internal access. I kept it at HIGH because once reachable, the compromise path is trivial, unauthenticated, and automatable, yielding a real shell on a sensitive appliance.
Why this verdict
- Start from 9.8, then cut for reachability: vendor scoring assumes any remote attacker can touch the service; in the real world this requires SSH exposure or prior network presence on a management/OT segment, so I take roughly -1.0 off the baseline.
- Cut again for blast radius: this CVE lands you on a single niche appliance, not an identity provider, VPN concentrator, or ubiquitous server tier. That narrows enterprise-wide impact even if the local device importance is high, worth about -0.4.
- Add back for ease and reliability: if reachable, the attack is basically
ssh user@hostwith a known trivial password and no exploit chain. That automatable simplicity keeps it solidly HIGH instead of drifting into MEDIUM.
Why not higher?
This is not a mass-market edge product with broad default internet exposure, and the exploit path depends on the management plane being reachable from the attacker's vantage point. Also, the CVE itself grants a user-level foothold on the appliance rather than proven root compromise by itself, so CRITICAL is too aggressive for fleet prioritization.
Why not lower?
MEDIUM would understate the danger where these boxes are reachable, because the attacker needs no prior credentials, no user interaction, and no exploit development. A hardcoded/trivial SSH login on a sensitive infrastructure appliance is still an immediate foothold and should not be treated as routine backlog noise.
What to do — in priority order.
- Restrict SSH to jump hosts — Limit TCP/22 on every SFX2100 to a dedicated management VLAN, VPN concentrator, or hardened bastion and deploy that control within 30 days. This directly removes the main amplifier for the CVE: attacker reachability.
- Rotate or disable the built-in account — Change the
useraccount password everywhere or lock/disable the account if operations allow, and finish that within 30 days. This neutralizes the credential-abuse path even if the firmware remains vulnerable. - Disable SSH where unnecessary — If the receiver can be managed through safer out-of-band paths or tightly-controlled alternatives, turn off SSH and complete the change within 30 days. Removing the service is cleaner than trying to detect every brute-force or reuse attempt.
- Alert on successful SSH logins — Forward auth logs if possible, or at minimum collect network telemetry for successful SSH sessions into the receiver fleet and stand up alerts within 30 days. On embedded appliances without EDR, auth and network logs are your best evidence source.
- A WAF does nothing here because the attack path is SSH, not HTTP.
- Changing only the Web GUI password does not fix a separate built-in SSH credential.
- Moving SSH to a non-standard port is not a control; scanners and targeted recon will still find it.
Crowdsourced verification payload.
Run this on the SFX2100 itself from a root shell or equivalent maintenance shell, because it needs read access to /etc/shadow. Save it as check-cve-2026-28777.sh, then run sudo bash check-cve-2026-28777.sh; it checks whether the user or usr account exists and whether its password still matches 12345.
#!/usr/bin/env bash
# check-cve-2026-28777.sh
# Determine whether IDC SFX2100 is vulnerable to CVE-2026-28777
# Exit codes: 0=PATCHED, 1=VULNERABLE, 2=UNKNOWN/runtime error, 3=UNKNOWN/insufficient privileges
set -u
if [[ ! -r /etc/shadow ]]; then
echo "UNKNOWN"
echo "Reason: need read access to /etc/shadow (run as root)." >&2
exit 3
fi
if ! command -v perl >/dev/null 2>&1; then
echo "UNKNOWN"
echo "Reason: perl is required to verify the shadow hash." >&2
exit 2
fi
check_account() {
local acct="$1"
local shadow_line hash result
shadow_line=$(grep -E "^${acct}:" /etc/shadow 2>/dev/null | head -n1 || true)
if [[ -z "$shadow_line" ]]; then
return 10
fi
hash=$(printf '%s' "$shadow_line" | cut -d: -f2)
# Locked / disabled / empty password states count as not vulnerable to this CVE path.
if [[ -z "$hash" || "$hash" == '!' || "$hash" == '*' || "$hash" == '!!' || "$hash" == '!*' ]]; then
return 20
fi
result=$(perl -e 'my ($pw,$h)=@ARGV; print crypt($pw,$h) eq $h ? "MATCH" : "NO";' '12345' "$hash" 2>/dev/null)
if [[ "$result" == "MATCH" ]]; then
return 30
fi
return 20
}
found_any=0
for acct in user usr; do
if grep -q -E "^${acct}:" /etc/passwd 2>/dev/null; then
found_any=1
check_account "$acct"
rc=$?
if [[ $rc -eq 30 ]]; then
echo "VULNERABLE"
echo "Reason: account '$acct' exists and still accepts password 12345." >&2
exit 1
fi
fi
done
if [[ $found_any -eq 0 ]]; then
echo "PATCHED"
echo "Reason: no 'user' or 'usr' account present." >&2
exit 0
fi
echo "PATCHED"
echo "Reason: built-in account exists but password no longer matches 12345, or account is locked." >&2
exit 0
If you remember one thing.
Sources
What defenders are saying.
Crowdsourced verification outputs.
Results submitted by users who ran the verification payload against their environment.