SECOPS

Cluster Cloudflare WAF events and draft tuned rule exceptions for review

On a schedule, pulls recent Cloudflare firewall events, clusters them by managed-rule ID and signature.

CategorySecOps
Enginesim
Difficultyintermediate
Triggerschedule
Steps5
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerDaily schedule fires
  • ActionFetch last 24h of WAF firewall eventsCloudflareCloudflare
  • LogicCluster by rule ID, URI, and ASN; rank false-positive likelihood
  • ActionDraft minimal skip-rule expressions and justificationsOpenAI
  • OutputOpen GitHub PR with drafted exceptions for reviewGitHubGitHub

What it does

Periodically reviews Cloudflare's managed WAF firewall events, groups them into behavioral clusters (same rule, path, and client fingerprint), then drafts surgical rule exceptions for the noisiest false-positive clusters. The exceptions land as a reviewable GitHub pull request, never as a live change.

When to use it

When a Cloudflare managed ruleset is generating recurring false positives on legitimate traffic (an internal API, a webhook callback, an admin path) and you want a tuned exception drafted for you each morning rather than hand-querying the events API and writing Terraform by hand.

How it works

  1. 1A daily schedule fires the workflow.
  2. 2It queries the Cloudflare GraphQL firewall-events API for the last 24h of `block`/`challenge` actions.
  3. 3A logic step clusters events by rule ID + URI pattern + ASN and ranks clusters by volume and likely-benign score.
  4. 4For each high-volume, low-risk cluster, an OpenAI step drafts a minimal skip-rule expression and a plain-English justification.
  5. 5The drafts are committed to a branch and opened as a GitHub PR tagged for the secops reviewer.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect CloudflareWorkers, Pages, R2, KV — the edge stack.
  2. 2
    Connect OpenAIModels, embeddings, files.
  3. 3
    Connect GitHubRepos, issues, pull requests, actions.
  4. 4
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  5. 5
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  6. 6
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.