ssrf-hunter

Detect server-side request forgery in URL fetchers, webhooks, and renderers using out-of-band verification.

1|Updated Aug 11, 2026
One-click install
npx skills add https://github.com/Maybe4a6f7365/agentic-bug-bounty-framework --skill ssrf-hunter-maybe4a6f7365
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ssrf-hunter
Source: https://github.com/Maybe4a6f7365/agentic-bug-bounty-framework/tree/main/skills/ssrf-hunter
Command: npx skills add https://github.com/Maybe4a6f7365/agentic-bug-bounty-framework --skill ssrf-hunter-maybe4a6f7365

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Server-side request forgery is hard to confirm because a reflected URL parameter alone proves nothing; this Skill provides a rigorous methodology to prove that a server actually fetches attacker-chosen destinations, using out-of-band oracles and negative controls so findings are report-ready rather than speculative. ## Core Features & Use Cases - Fetcher Surface Inventory: Systematically enumerate SSRF sinks including webhooks, import-from-URL, link previews, PDF/screenshot renderers, media proxies, and OIDC/SAML metadata discovery. - Out-of-Band Verification: Confirm server-initiated requests with interactsh, Burp Collaborator, or self-hosted DNS/HTTP collectors using unique per-test canaries. - Bypass and Escalation Recipes: Test allowlist bypasses (redirect chains, DNS rebinding, parser confusion) and prove read-only reach to AWS, GCP, and Azure metadata endpoints without harvesting credentials. - Use Case: While testing a bug-bounty target's link-preview feature, point it at a unique collector subdomain, verify the inbound request originates from the target's egress IP, then demonstrate redacted reach to the cloud metadata endpoint for a reportable finding. ## Quick Start Use the ssrf-hunter skill to test the target's URL preview endpoint for server-side request forgery with an out-of-band canary.

Frequently Asked Questions about ssrf-hunter

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I test for SSRF vulnerabilities in a web application?

Identify features that fetch user-supplied URLs such as webhooks, importers, and link previews, then point them at a unique out-of-band collector subdomain. SSRF is confirmed only when your collector records a request from the target's egress IP or the response returns internal-only bytes.

What tools can I use as an out-of-band SSRF oracle?

Burp Collaborator is the primary option, with interactsh as the fallback and default when using Caido, which lacks a Collaborator equivalent. Self-hosted DNS/HTTP callback servers or pipedream and webhook.site offer progressively less control.

How do I bypass SSRF allowlists and IP filters?

Test redirect chains from allowlisted hosts, DNS rebinding where validation and fetch resolve differently, alternate IP representations like decimal, hex, or IPv6, and parser confusion between the validating URL parser and the fetching client.

Can SSRF reach cloud metadata endpoints like AWS 169.254.169.254?

Yes, but only prove reach with a read-only request and redact results; never pull real credentials from iam/security-credentials. Note that AWS IMDSv2 requires a token via PUT, so a failing plain GET is a control, not a finding.

Why is a reflected URL parameter not proof of SSRF?

A reflected or accepted url parameter shows only that input was received, not that the server made a request. Without an out-of-band collector hit or internal-only response bytes, there is no receiver oracle and the finding is theoretical.

When should I stop escalating an SSRF finding?

Stop at proof-of-reach: one or two redacted samples confirming internal or metadata access. Never port-scan internal ranges at volume, exfiltrate live credentials, or claim RCE without a demonstrated chain.