sast-ssrf

Name: sast-ssrf
Author: utkusen

37installs56forks

How to install sast-ssrf?

utkusen/sast-skills/sast-ssrf

$npx -y skills add utkusen/sast-skills --skill sast-ssrf

Installs into the current project.

Summary

Detect Server-Side Request Forgery (SSRF) vulnerabilities in a codebase using a three-phase approach: recon (find outbound call sites), batched verify (trace user input to destinations in parallel subagents, 3 sites each), and merge (consolidate batch results). Requires sast/arch

Files · 1

View on GitHub

SKILL.md

1	# Server-Side Request Forgery (SSRF) Detection
2
3	You are performing a focused security assessment to find SSRF vulnerabilities in a codebase. This skill uses a three-phase approach with subagents: recon (find all places that make outbound TCP, DNS, or HTTP requests), batched verify (trace whether user-supplied input reaches those call sites, in parallel batches of 3), and merge (consolidate batch reports into one file).
4
5	Prerequisites: `sast/architecture.md` must exist. Run the analysis skill first if it doesn't.
6
7	---
8
9	## What is SSRF
10
11	SSRF occurs when an attacker can cause the server to make outbound network requests to an arbitrary destination — including internal services, cloud metadata endpoints, or other external targets — by supplying or influencing the URL, hostname, IP, or port used in a server-side request.
12
13	The core pattern: unvalidated, user-controlled input reaches the destination argument of an outbound network call.
14
15	### What SSRF IS
16
17	- HTTP client calls where the URL or host is built from user input: `requests.get(user_url)`
18	- Fetching a resource whose location is provided by the client: `fetch(req.body.webhook_url)`
19	- DNS lookups on a hostname supplied by the user: `dns.lookup(req.query.host)`
20	- Raw TCP connections to a host/port derived from user input: `socket.connect((user_host, user_port))`
21	- File-fetching functions used with HTTP/FTP URLs from user input: `file_get_contents($user_url)`
22	- URL redirectors that forward to a user-supplied destination without validation
23	- Webhooks, import-from-URL, screenshot services, PDF renderers, image proxies — any feature that fetches a remote resource on behalf of the user
24
25	### What SSRF is NOT
26
27	Do not flag these:
28
29	- Open redirects: Redirecting the browser (HTTP 302) to a user-supplied URL — that's a client-side redirect, not a server-side request
30	- XSS via URL: Rendering a user-supplied URL in an `<a>` tag without escaping — that's XSS
31	- IDOR: Accessing another user's data by changing an object ID — separate vulnerability class
32	- Hardcoded outbound calls: HTTP requests to fixed, fully hardcoded URLs with no user influence — not SSRF
33
34	### Patterns That Prevent SSRF
35
36	When you see these patterns, the code is likely not vulnerable:
37
38	1. Strict allowlist of permitted destinations
39	```python
40	ALLOWED_HOSTS = {"api.example.com", "cdn.example.com"}
41	parsed = urlparse(user_url)
42	if parsed.hostname not in ALLOWED_HOSTS:
43	raise ValueError("Destination not allowed")
44	requests.get(user_url)
45	```
46
47	2. Allowlist of permitted URL prefixes / schemes
48	```python
49	ALLOWED_PREFIXES = ["https://api.example.com/", "https://cdn.example.com/"]
50	if not any(user_url.startswith(p) for p in ALLOWED_PREFIXES):
51	abort(400)
52	requests.get(user_url)
53	```
54
55	3. No user influence on the destination
56	```python
57	# Destination fully hardcoded — no user input involved
58	response = requests.get("https://api.thirdparty.com/data")
59	```
60
61	> Note: IP blocklists (blocking 169.254.0.0/16, 10.0.0.0/8, etc.) are not sufficient protection — they can be bypassed via DNS rebinding, URL encoding, IPv6 notation, decimal IP representation, or redirect chains. Do not treat a blocklist as making a site safe; classify it as Likely Vulnerable.
62
63	---
64
65	## Vulnerable vs. Secure Examples
66
67	### Python — requests
68
69	```python
70	# VULNERABLE: URL fully controlled by user
71	@app.route('/fetch')
72	def fetch():
73	url = request.args.get('url')
74	response = requests.get(url)
75	return response.text
76
77	# SECURE: strict allowlist on destination host
78	ALLOWED = {"api.example.com"}
79	@app.route('/fetch')
80	def fetch():
81	url = request.args.get('url')
82	if urlparse(url).hostname not in ALLOWED:
83	abort(403)
84	response = requests.get(url)
85	return response.text
86	```
87
88	### Python — urllib
89
90	```python
91	# VULNERABLE: user controls the URL passed to urlopen
92	def preview(request):
93	target = request.GET.get('target')
94	data = urllib.request.urlopen(target).read()
95	return HttpResponse(data)
96
97	# SECURE: only allow https scheme to a hardcoded host
98	def preview(request):
99	target = request.GET.get('target')
100	parsed = urlparse(target)
101	if parsed.scheme != 'https' or parsed.hostname != 'media.example.com':
102	return HttpResponse(status=400)
103	data = urllib.request.urlopen(target).read()
104	return HttpResponse(data)
105	```
106
107	### Node.js — fetch / axios
108
109	```javascript
110	// VULNERABLE: webhook URL comes directly from request body
111	app.post('/webhook/test', async (req, res) => {
112	const { url } = req.body;
113	const res

People who liked this also viewed

6 picks

Type

sast-ssrf

37installs56forks

View on GitHub

How to install sast-ssrf?

utkusen/sast-skills/sast-ssrf

$npx -y skills add utkusen/sast-skills --skill sast-ssrf

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Server-Side Request Forgery (SSRF) Detection
2
3	You are performing a focused security assessment to find SSRF vulnerabilities in a codebase. This skill uses a three-phase approach with subagents: recon (find all places that make outbound TCP, DNS, or HTTP requests), batched verify (trace whether user-supplied input reaches those call sites, in parallel batches of 3), and merge (consolidate batch reports into one file).
4
5	Prerequisites: `sast/architecture.md` must exist. Run the analysis skill first if it doesn't.
6
7	---
8
9	## What is SSRF
10
11	SSRF occurs when an attacker can cause the server to make outbound network requests to an arbitrary destination — including internal services, cloud metadata endpoints, or other external targets — by supplying or influencing the URL, hostname, IP, or port used in a server-side request.
12
13	The core pattern: unvalidated, user-controlled input reaches the destination argument of an outbound network call.
14
15	### What SSRF IS
16
17	- HTTP client calls where the URL or host is built from user input: `requests.get(user_url)`
18	- Fetching a resource whose location is provided by the client: `fetch(req.body.webhook_url)`
19	- DNS lookups on a hostname supplied by the user: `dns.lookup(req.query.host)`
20	- Raw TCP connections to a host/port derived from user input: `socket.connect((user_host, user_port))`
21	- File-fetching functions used with HTTP/FTP URLs from user input: `file_get_contents($user_url)`
22	- URL redirectors that forward to a user-supplied destination without validation
23	- Webhooks, import-from-URL, screenshot services, PDF renderers, image proxies — any feature that fetches a remote resource on behalf of the user
24
25	### What SSRF is NOT
26
27	Do not flag these:
28
29	- Open redirects: Redirecting the browser (HTTP 302) to a user-supplied URL — that's a client-side redirect, not a server-side request
30	- XSS via URL: Rendering a user-supplied URL in an `<a>` tag without escaping — that's XSS
31	- IDOR: Accessing another user's data by changing an object ID — separate vulnerability class
32	- Hardcoded outbound calls: HTTP requests to fixed, fully hardcoded URLs with no user influence — not SSRF
33
34	### Patterns That Prevent SSRF
35
36	When you see these patterns, the code is likely not vulnerable:
37
38	1. Strict allowlist of permitted destinations
39	```python
40	ALLOWED_HOSTS = {"api.example.com", "cdn.example.com"}
41	parsed = urlparse(user_url)
42	if parsed.hostname not in ALLOWED_HOSTS:
43	raise ValueError("Destination not allowed")
44	requests.get(user_url)
45	```
46
47	2. Allowlist of permitted URL prefixes / schemes
48	```python
49	ALLOWED_PREFIXES = ["https://api.example.com/", "https://cdn.example.com/"]
50	if not any(user_url.startswith(p) for p in ALLOWED_PREFIXES):
51	abort(400)
52	requests.get(user_url)
53	```
54
55	3. No user influence on the destination
56	```python
57	# Destination fully hardcoded — no user input involved
58	response = requests.get("https://api.thirdparty.com/data")
59	```
60
61	> Note: IP blocklists (blocking 169.254.0.0/16, 10.0.0.0/8, etc.) are not sufficient protection — they can be bypassed via DNS rebinding, URL encoding, IPv6 notation, decimal IP representation, or redirect chains. Do not treat a blocklist as making a site safe; classify it as Likely Vulnerable.
62
63	---
64
65	## Vulnerable vs. Secure Examples
66
67	### Python — requests
68
69	```python
70	# VULNERABLE: URL fully controlled by user
71	@app.route('/fetch')
72	def fetch():
73	url = request.args.get('url')
74	response = requests.get(url)
75	return response.text
76
77	# SECURE: strict allowlist on destination host
78	ALLOWED = {"api.example.com"}
79	@app.route('/fetch')
80	def fetch():
81	url = request.args.get('url')
82	if urlparse(url).hostname not in ALLOWED:
83	abort(403)
84	response = requests.get(url)
85	return response.text
86	```
87
88	### Python — urllib
89
90	```python
91	# VULNERABLE: user controls the URL passed to urlopen
92	def preview(request):
93	target = request.GET.get('target')
94	data = urllib.request.urlopen(target).read()
95	return HttpResponse(data)
96
97	# SECURE: only allow https scheme to a hardcoded host
98	def preview(request):
99	target = request.GET.get('target')
100	parsed = urlparse(target)
101	if parsed.scheme != 'https' or parsed.hostname != 'media.example.com':
102	return HttpResponse(status=400)
103	data = urllib.request.urlopen(target).read()
104	return HttpResponse(data)
105	```
106
107	### Node.js — fetch / axios
108
109	```javascript
110	// VULNERABLE: webhook URL comes directly from request body
111	app.post('/webhook/test', async (req, res) => {
112	const { url } = req.body;
113	const res

People who liked this also viewed

6 picks

Type