OneFirewall vs Cloudflare WAF: block attackers before they reach your origin.
Cloudflare's WAF inspects HTTP requests at the edge. OneFirewall tells it which source IPs are already attacking other organisations, through the Cloudflare API, so known attackers are refused at the first request on every plan.
WAF & cloud edge · Last reviewed: 20 November 2026
Cloudflare and OneFirewall are complementary, focused here on the WAF layer. Cloudflare decides whether a request looks malicious: managed rules, custom rules, rate limiting and bot scores. OneFirewall decides whether the client is a known attacker, based on what Alliance members saw on firewalls, VPNs and servers, not just on websites. The WCF Agent writes that decision into your Cloudflare configuration automatically.
A global reverse proxy with a programmable WAF
Cloudflare sits in front of your site as an anycast reverse proxy. Every HTTP request passes through its edge, where the WAF evaluates managed rulesets (Cloudflare and OWASP core), your custom rules written in the Rules language, rate-limiting rules and, on higher plans, Bot Management scores. Non-matching traffic is proxied to your origin.
For known bad clients, Cloudflare offers IP Access rules and Lists that custom rules can reference. Both can be managed through the Cloudflare API, which is how an external intelligence source keeps them current. Feature depth and list limits depend on your plan.
- Huge edge network with DDoS mitigation and caching
- WAF decisions are based mainly on what each request looks like
- Bot Management and advanced features vary by plan
- IP-based blocks are managed by you, by hand or via the API
The WCF Agent manages Cloudflare through its API
Cloudflare is an agent-managed integration: you give the WCF Agent API credentials, and it keeps Cloudflare in step with the Alliance feed in both directions.
Create API credentials
In Cloudflare, create credentials for the account and zones to protect, and add them to the cloudflare section of the WCF Agent configuration.
Set the threshold and interval
In Install Agent, choose the IoC type, the Crime Score threshold and the sync interval. Only IPs above the threshold are pushed.
Automatic sync
On each cycle the agent adds newly confirmed attackers to your Cloudflare block configuration and removes those whose score has decayed. Blocked requests never reach your origin.
Feedback from the edge
With the ids side enabled, the agent reads Cloudflare firewall events back into OneFirewall, so attacks on your site feed your Defence Center and, if you contribute, the Alliance.
What the configuration looks like
// WCF Agent config.json
"ips": { // push: keep the zone block list in sync
"cloudflare": {
"active": true,
"cloudflare_x_auth_email": "secops@example.com",
"cloudflare_x_auth_key": "********"
}
},
"ids": { // read: firewall events back into OneFirewall
"cloudflare": { "active": true, "cloudflare_x_auth_email": "...", "cloudflare_x_auth_key": "..." }
}
In our SaaS case study, the customer was on Cloudflare's free plan with no advanced DDoS or bot protection. OneFirewall layered enforcement on top and stopped 35,000 daily attacks within 24 hours, without changing CDN.
Cloudflare and OneFirewall, capability by capability
| Capability | Cloudflare WAF | OneFirewall |
|---|---|---|
| Question it answers | Does this request look malicious? | Is this client a known attacker right now? |
| Signal | Request content, managed rules, rate limits, bot score | Attacks reported by 290+ organisations on firewalls, VPNs, servers and web apps |
| Coverage of non-web attacks | Sees only HTTP(S) traffic to proxied hostnames | Knows IPs from SSH brute force, port scans and exploit attempts seen elsewhere |
| Plan dependency | Advanced bot and WAF features depend on plan | Same Alliance feed on any plan, through the API |
| List upkeep | IP rules and lists maintained by you | Added and removed automatically by Crime Score and decay |
| Beyond the edge | Protects proxied web properties | The same list protects your firewalls, cloud load balancers and SIEM |
What OneFirewall adds to Cloudflare
Context Cloudflare can't see
An IP brute-forcing a member's VPN at 9:00 often hits web logins at 9:30. The Alliance sees the first attack, so Cloudflare can refuse the second.
Advanced blocking on any plan
Teams on free or Pro plans get curated, maintained attacker blocking without moving to an enterprise tier.
One policy across edge and perimeter
The same scored list blocks at Cloudflare, at your firewalls and in Cloud Armor or AWS WAF, so an attacker can't route around one control.
- Volumetric DDoS mitigation and caching
- Request-level inspection: SQL injection, XSS and OWASP rules
- Rate limiting, bot scoring and challenges
- TLS termination and edge performance
OneFirewall doesn't proxy or inspect your web traffic. It improves the decisions Cloudflare makes at the edge.
OneFirewall and Cloudflare: FAQ
Does OneFirewall replace Cloudflare?
No. Cloudflare keeps proxying, caching and inspecting your traffic. OneFirewall keeps its IP-based blocking current with Alliance-confirmed attackers.
How does OneFirewall connect to Cloudflare?
Through the Cloudflare API. You add API credentials to the WCF Agent configuration and it manages the block configuration on each sync cycle.
Does it work on Cloudflare's free plan?
Yes. Our published case study started on Cloudflare's free plan. Available list sizes depend on your plan, which the Crime Score threshold helps you stay within.
Can Cloudflare events flow back into OneFirewall?
Yes. The agent's ids module reads Cloudflare firewall events so they appear in your OneFirewall dashboards and can be contributed to the Alliance.
