Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessConnect as authorized MCP client
-
DeliveryInvoke get_reference tool
-
ExploitSupply traversing Name argument
-
ExecutionServer appends .md and resolves outside references
-
ImpactRead Markdown file outside skill directory
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A01:2025 - Broken Access Control
ASVS 5.0
V5 - File Handling Path canonicalization, traversal prevention
MITRE ATT&CK
2 techniques across 2 tactics: Discovery, Collection The technique list with ATT&CK links and the kill-chain view is part of the Basic plan and up. |
| Exploitation Conditions | Exploitation requires the attacker to be a connected MCP client able to invoke the get_reference tool of the designer-mcp MCP Interface (PR:L - authenticated per the CVSS 4.0 vector). The vulnerable get_reference function in skills/motion-designer/references concatenated the Name argument directly into a filesystem path and appended a '.md' extension, so traversal was constrained to Markdown-suffixed files, and impact is limited to low confidentiality (file read) with no write or availability effect. Limiting factors: the caller must hold MCP-client access to the tool, the read is restricted to '.md'-extension targets, and deployments that run the server locally over stdio rather than exposing it on a network materially reduce real-world reachability. Affected: designer-mcp up to 2.1.0; fixed by commit c8c2951dcbed9b1710a745993f028317f10a3f10 (no version numbers due to rolling releases). |
| Impact | This is a moderate-severity, low-urgency issue rather than a top priority. The supplied CVSS 4.0 vector (5.3) describes a network-accessible flaw with low attack complexity but requiring privileges (PR:L) and yielding only low confidentiality impact (VC:L) with no integrity or availability impact (VI:N/VA:N). The E:P exploit-maturity modifier reflects a proof-of-concept, consistent with the input noting a PoC reference that is not verified as working exploit code. There is no CISA KEV listing and no EPSS score in the provided data, so exploitation-probability cannot be quantified; the vuln.today decision label is 'this_month', not act_now or emergency. The SSVC-style signals therefore agree on limited urgency: a confirmed but unweaponized read-only path traversal in a niche MCP server. One caveat to verify: the description and CVSS both assert remote/network exploitation (AV:N), but designer-mcp is a Model Context Protocol server that is frequently run locally over stdio (the vendor's own regression test 'drives the real server over stdio'), so the true attack vector depends on whether a given deployment exposes the MCP interface over a network - verify against the actual deployment model. |
| Exploit ScenarioDerived | An attacker who can act as a connected MCP client invokes the get_reference tool and supplies a Name argument containing directory-traversal sequences; because the pre-patch code concatenated the argument directly into a path under the references directory (appending a .md extension), the server resolves and returns a Markdown file outside the intended skill directory. A public PoC reference exists (a GitHub issue), but it is not verified as working exploit code. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 25, 2026
A path traversal flaw in the designer-mcp MCP server (versions up to 2.1.0) lets a caller of the get_reference tool read files outside the bundled references directory by supplying a crafted reference name. Because the argument was concatenated directly into a path with a '.md' extension appended, exposure is limited to Markdown files and to low confidentiality impact - no write or availability effect. Exploitation requires the attacker to already hold access as a connected MCP client invoking the tool, and a public issue report (#14) referencing the flaw exists but is not verified working exploit code; there is no confirmation of active exploitation. The project ships rolling releases, so version-based guidance is unavailable and remediation is by commit c8c2951dcbed9b1710a745993f028317f10a3f10.
RemediationAI
Apply the vendor fix commit c8c2951dcbed9b1710a745993f028317f10a3f10 (https://github.com/1999AZZAR/designer-mcp/commit/c8c2951dcbed9b1710a745993f028317f10a3f10), which routes get_reference and get_ellis_ui_template through resolveWithinRoot() in src/envelope.ts; because the project ships rolling releases, pinning to that commit or any build that includes it is the only reliable version guidance. Per the vendor advisory (https://vuldb.com/vuln/426121) applying the patch is advised; no separate workaround is published. Where immediate patching is not possible, compensating controls are to stop exposing get_reference to untrusted MCP clients - removing or gating the tool prevents the traversal but also removes reference-documentation access for legitimate users - or to run the server locally over stdio rather than exposing it as a network-reachable MCP endpoint, which reduces reachability at the cost of remote client convenience. Note that after the fix, pre_flight_scan and the logo_* tools still accept caller-supplied paths by design, so clients granted those tools retain file-read capability within the caller's own permissions.
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
designer-mcp is a Node/TypeScript MCP (Model Context Protocol) server that exposes tools to LLM clients; the vulnerable component is the MCP Interface's get_reference handler, which serves bundled documentation from skills/motion-designer/references. Root cause is CWE-22 (Improper Limitation of a Pathname to a Restricted Directory): the caller-supplied Name argument was joined straight into the filesystem path and suffixed with '.md', so traversal sequences in Name could resolve to a Markdown file outside the intended skill directory. The fix introduces resolveWithinRoot() in src/envelope.ts, applying three independent checks - a lexical check rejecting separators, '..', absolute paths and NUL bytes; resolution against the realpath of the trusted root so a symlinked root cannot widen the boundary; and a final relative() check confirming the resolved target (and, if present, its realpath) still sits under that base. The same helper is applied to get_ellis_ui_template; the commit's documentation notes that pre_flight_scan and the logo_* tools deliberately continue to accept caller-chosen paths because supplying a project directory or artwork file is their intended function.
Know when Designer Mcp 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:L/AT:N/PR:L/UI:N/VC:L/VI:N/VA:N/SC:N/SI:N/SA:N/E:P
Caller must be an authorized MCP client invoking the tool (PR:L); traversal is simple (AC:L); only Markdown files can be read, giving low confidentiality (C:L) and no integrity or availability impact.
Primary rating from Vendor (VulDB).
CVSS VectorVendor: VulDB
Lifecycle Timeline
3DescriptionCVE.org
A vulnerability was detected in 1999AZZAR designer-mcp up to 2.1.0. This affects the function get_reference of the file skills/motion-designer/references/motion-dev-guidelines.md of the component MCP Interface. The manipulation of the argument Name results in path traversal. It is possible to launch the attack remotely. The exploit is now public and may be used. This product takes the approach of rolling releases to provide continious delivery. Therefore, version details for affected and updated releases are not available. The patch is identified as c8c2951dcbed9b1710a745993f028317f10a3f10. Applying a patch is advised to resolve this issue.
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
Path traversal exploits occur when applications use user-controlled input to construct file system paths without proper validation.
EUVD-2026-109354
GHSA-p45h-7j97-qg29
NVD
CVE.org
EPSS
VulDB
Exploit-DB