Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAuthenticate as project manager with connector.write
-
ExploitSet connector URL to IPv6 6to4/Teredo wrapping private IPv4
-
ExecutionBypass isPrivateIp guard, server fetches internal target
-
ImpactRead metadata/internal service response
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A01:2025 - Broken Access Control
ASVS 5.0
V4 - API and Web Service URL allowlisting, SSRF prevention
MITRE ATT&CK
2 techniques across 2 tactics: Initial Access, Credential Access The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | The attacker must be an authenticated user holding the project.connector.write permission (project-manager role) on a Suna instance running an affected version (0.10.7 up to but not including 0.13.52). They must supply a connector base_url, OpenAPI, Postman, or MCP URL encoded as an IPv6 6to4 (2002::/16) or Teredo (2001:0::/32) address whose embedded IPv4 points at an internal/RFC1918 host or the 169.254.169.254 cloud metadata endpoint; plain private IPv4 is already blocked by isPrivateIp. Limiting factors: this is not remotely unauthenticated - it needs insider connector-write privilege, requires crafting a valid 6to4/Teredo wrapper (reflected as AC:H/AT:P), only works against server deployments that reach the target internal network, and is fully remediated in v0.13.52. |
| Impact | This is a genuinely low-priority issue despite being an SSRF with cloud-metadata reach. The vendor-supplied CVSS 4.0 vector (AV:N/AC:H/AT:P/PR:L/UI:N) scores only 2.3, and that low score is credible rather than a case of 'high CVSS, low real risk' - the opposite problem. The attack requires an already-trusted insider: a user holding the project.connector.write permission (PR:L, so authenticated), plus a crafted non-trivial IPv6 encoding (AC:H and AT:P reflect the present attack requirement of building valid 6to4/Teredo wrappers around private IPv4 targets). Impact is capped at partial read of responses (VC:L/VI:L, VA:N) and partial subsequent-system impact (SC:L/SI:L) - consistent with reading internal service and metadata endpoint responses but not full credential theft or write. A public POC exists (VulnCheck advisory plus a hackmd technical write-up), which raises practical exploitability, but there is NO EPSS score provided and NO CISA KEV listing, so there is no evidence of active or widespread exploitation - the signal is 'exploit code available, no observed exploitation.' The POC and low CVSS are consistent with each other: an authenticated-only, config-dependent SSRF bypass. Note one missing signal explicitly: EPSS is absent from the input, so widespread-exploitation probability cannot be quantified here. |
| Exploit ScenarioDerived | A user granted project-manager rights (project.connector.write) in a self-hosted Suna instance configures a connector's base_url, OpenAPI, Postman, or MCP URL to an IPv6 6to4 (e.g. 2002:a9fe:a9fe::) or Teredo address that embeds 169.254.169.254 or an RFC1918 host. The isPrivateIp guard fails to decode the embedded private IPv4, Suna fetches the URL server-side, and the attacker reads back cloud metadata or internal service responses. A public proof-of-concept and technical write-up exist, but exploitation requires the attacker to already hold the connector-write privilege. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 10, 2026
Server-side request forgery in Kortix Suna 0.10.7 through 0.13.51 allows an authenticated user holding the project.connector.write permission to bypass the isPrivateIp egress guard by encoding connector base_url, OpenAPI, Postman, or MCP endpoints as IPv6 6to4 (2002::/16) or Teredo (2001:0::/32) addresses whose embedded IPv4 points at an internal host or the 169.254.169.254 cloud metadata endpoint. Exploitation reaches internal-only services and metadata and returns the fetched responses to the attacker, but impact is capped at low confidentiality and integrity with no availability loss and no scope escape beyond the connector fetch path. Publicly available exploit code exists (a technical writeup linked from the CVE record), and the issue is fully remediated in v0.13.52; there is no confirmed active exploitation (CISA KEV) for this CVE.
RemediationAI
Vendor-released patch: upgrade to Kortix Suna v0.13.52 or later (release: https://github.com/kortix-ai/suna/releases/tag/v0.13.52; fix commit: https://github.com/kortix-ai/suna/commit/9c949e4d876cd5acf7b23b0a9ee49ca5c53f1332), which extends isPrivateIp to cover 6to4/Teredo/IETF-reserved ranges and also pins egress connections to validated DNS answers. If immediate upgrade is not possible, apply layered compensating controls: restrict the project.connector.write permission to a small set of trusted administrators and audit connector records for IPv6 literal URLs, especially 2002::/16 and 2001:0::/32 forms, since those are the only addresses that bypass the guard (trade-off: legitimate 6to4/Teredo connector endpoints are effectively unusable until patched, which is acceptable because such configurations are rare for SaaS-style connectors); enforce egress network policy or a forward proxy on the Suna API host that denies RFC1918, link-local, and 169.254.169.254 destinations (trade-off: any connector that legitimately targets on-premise internal services will break and must be re-scoped or allowlisted explicitly); and on cloud hosts require IMDSv2 with a low hop limit or disable the metadata service entirely so that even a successful fetch cannot return credentials (trade-off: instance IAM flows that rely on IMDSv1 will stop working). Note that the local writeup at https://hackmd.io/@haind/suna-ssrf-6to4-guard-bypass demonstrates the technique and should be used to build detection signatures for outbound requests to 6to4/Teredo-encoded private destinations while patching proceeds.
Affected VersionsENISA
Affected Packages
Source: OSV (Open Source Vulnerabilities)
Affects your stack
Affects 1 package in euvd
Checking this CVE against your own stacks and servers is part of the Basic plan and up.
Technical ContextAI · show details
The vulnerability is a classic CWE-918 server-side request forgery rooted in incomplete IP-address normalization inside Suna's egress guard (apps/api/src/shared/ssrf-guard.ts, referenced lines 98-113). The isPrivateIp check correctly rejects plain RFC1918 IPv4 literals, but it does not decode IPv6 transition/tunneling formats: a 6to4 address in 2002::/16 embeds a 32-bit IPv4 address in bits 17-48, and a Teredo address in 2001:0::/32 embeds the client's IPv4 address in the final 32 bits XOR-obfuscated with 0xFF. When a connector URL is supplied as an IPv6 literal such as 2002:a9fe:a9fe:: (6to4-wrapping 169.254.169.254), the guard sees an unfamiliar IPv6 address and permits the fetch, while the host network stack routes the connection to the embedded private IPv4 target. Because the guard is applied to user-controlled connector configuration fields (base_url, OpenAPI spec URL, Postman collection URL, MCP server URL), any principal with project.connector.write can turn the Suna API server into an internal-network proxy. The upstream fix commit 9c949e4d876cd5acf7b23b0a9ee49ca5c53f1332 extends isPrivateIp to recognize 6to4 (including 2002:7f00:1:: and 2002:0a00:0001::) and Teredo ranges plus IETF protocol-assignment and 6to4 relay-anycast space, and the same change set hardens adjacent SSRF surface: safeEgressFetch now pins the connection to the DNS answer it validated (preventing a second lookup from rebinding to 169.254.169.254), preserves the original Host header, refuses cross-host redirects on bearer-authenticated downloads, and canonicalizes proxy request paths once before authorization decisions.
Know when Suna gets its next vulnerability
- Watch products and get an email when something new is published
- Check this CVE against your own stack and servers
- Exploit and PoC references, threat intelligence from 22 sources
No credit card · 7-day full trial
Severity by source
CVSS:4.0/AV:N/AC:H/AT:P/PR:L/UI:N/VC:L/VI:L/VA:N/SC:L/SI:L/SA:N
Authenticated connector-write privilege gives PR:L, crafting a valid 6to4/Teredo wrapper gives AC:H, SSRF reaching other systems is a scope change (S:C) with partial read/write impact and no availability effect.
Primary rating from Vendor (VulnCheck).
CVSS VectorVendor: VulnCheck
DescriptionCVE.org
Kortix Suna 0.10.7 before 0.13.52 contains a server-side request forgery vulnerability that allows project managers to bypass the isPrivateIp guard by supplying IPv6 6to4 or Teredo addresses that embed private IPv4 destinations. Attackers holding project.connector.write can set connector base_url, OpenAPI, Postman, or MCP URLs to reach internal services and cloud metadata endpoints and read responses.
Threat Intelligence Overview
How this decision was reached
The engine reads each signal below and applies a fixed list of rules from top to bottom. The first rule that matches sets the recommended action – like a statement, every line is shown.
What every action level means
SSVC AssessmentENISA
Priority Score
Attack Technique
Server-Side Request Forgery exploits applications that fetch remote resources based on user-supplied URLs.
We are tasked to produce a JSON analysis for CVE-2026-66027 based on the provided data. The analysis must synthesize all
Reflected cross-site scripting in Suna's authentication endpoint allows remote attackers to execute arbitrary JavaScript
EUVD-2026-96339
GHSA-f37m-qrrx-ww7w
NVD
CVE.org
EPSS
Exploit-DB