Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessAuthenticated attacker configures custom API tool with attacker-controlled URL
-
DeliverySocial-engineer privileged user to run the tool
-
ExploitServer sends victim's auth token to attacker URL
-
ExecutionCapture leaked Splunk platform token
-
ImpactReplay token to impersonate victim
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 | Exploitation requires a multi-user configuration: (1) an attacker must have the ability to create/configure a custom API tool in Splunk MCP Server and set its target URL to an endpoint they control, and (2) a separate victim user who holds a role containing the mcp_tool_execute capability must actively run that attacker-configured custom API tool. Only then is the victim's Splunk platform authentication token sent to the attacker URL. Limiting factors that reduce real-world risk: the victim's required mcp_tool_execute privilege, the mandatory user interaction (the victim must choose to run the tool), the attacker's need for tool-configuration privileges, and the dependency on the custom-tool feature being in use - this does not affect default deployments that do not use custom API tools. |
| Impact | This is a genuine but moderate-priority issue, not a drop-everything emergency. The CVSS 3.1 base score of 5.3 (Medium) is held down by two strong limiting metrics: AC:H (the attack depends on a specific multi-user workflow coming together) and PR:H combined with UI:R (the attacker must already be a privileged tool author AND a separate, higher-value victim must voluntarily run the malicious tool). The upside for an attacker is significant - C:H reflects theft of a live Splunk platform authentication token that lets them impersonate the victim. No EPSS score was provided in the input, so exploitation probability cannot be quantified. The vulnerability is NOT in CISA KEV and no POC was supplied, so this is no public exploit identified at time of analysis. The signals are internally consistent rather than conflicting: a network-reachable SSRF-style token exfiltration (CWE-918) that is gated behind privilege and social-engineering requirements, which is exactly why a high-impact confidentiality flaw lands at only Medium severity. The primary real-world risk is insider/multi-tenant abuse rather than mass exploitation. Note: EPSS is absent and should be checked before final prioritization. |
| Exploit ScenarioDerived | A user with rights to create custom API tools configures a tool pointing at a URL they control. A second, more privileged user holding the mcp_tool_execute capability runs that tool, causing Splunk MCP Server to transmit the running user's Splunk platform authentication token to the attacker-controlled URL. The attacker captures the token and replays it to read data and perform actions as the victim. No public exploit code was identified at time of analysis. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 07, 2026
Splunk MCP Server versions below 1.2.1 transmit the Splunk platform authentication token of the user who executes a custom API tool to whatever URL that tool is configured to call, so an attacker who can author a custom tool pointing at a server they control can capture the victim's token and then act as that user. Exploitation is authenticated and requires user interaction: the attacker needs privileges to create or configure a custom API tool, and a separate victim holding a role with the mcp_tool_execute capability must actively run that tool (CVSS:3.1/AV:N/AC:H/PR:H/UI:R/S:U/C:H/I:L/A:L). Real-world risk is moderate - it does not affect default deployments that make no use of custom API tools, and no public exploit has been identified at time of analysis. A vendor patch is available in Splunk MCP Server 1.2.1.
RemediationAI
Upgrade to Splunk MCP Server 1.2.1 or later - the vendor-released fix is 1.2.1, documented in Splunk advisory SVD-2026-1004 (https://advisory.splunk.com/advisories/SVD-2026-1004). Until the upgrade is applied, the most effective compensating control is to eliminate the attack path rather than harden the request: disable or remove custom API tools that point at endpoints not fully controlled by the Splunk deployment, and review every existing custom tool definition for untrusted or third-party target URLs, since any such tool is a token-exfiltration channel when run. Restrict the ability to create and configure custom API tools to a small, trusted administrative set, and audit who currently holds that capability - the trade-off is reduced self-service for teams that legitimately author their own tools, and existing attacker-planted tool definitions will not be found by permission changes alone, so an inventory pass is required. Narrowing the mcp_tool_execute capability to only the roles that genuinely need it limits the pool of victims but does not remove the flaw, because any user granted that capability can still trigger token forwarding. If custom API tools cannot be disabled, treat the Splunk tokens of all users holding mcp_tool_execute as potentially exposed, rotate them after remediation, and monitor Splunk audit and authentication logs for tool executions followed by anomalous activity from the affected accounts - this is detection after the fact and does not prevent token capture. Note that the vendor states a patch is available; no workaround that fully closes the issue short of upgrading or disabling custom API tools has been published.
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 root cause is CWE-918 (Server-Side Request Forgery) manifesting in Splunk MCP Server, the Model Context Protocol bridge that exposes Splunk platform data and actions to AI agents and clients through callable tools. Custom API tools are user-defined entries whose configuration includes a target URL; when a user invokes such a tool, the MCP Server makes an outbound request to that URL and, per the advisory, includes the invoking user's Splunk platform authentication token. The server does not appear to validate that the destination URL is trusted or that token forwarding is appropriate for user-supplied endpoints, so an attacker-controlled URL receives a credential belonging to the victim rather than the attacker. The CREST/SSRF pattern is bounded by the CVSS vector: AC:H reflects the dependency on a second user with tool-authoring rights and on the custom-tool feature being enabled and in use; PR:H reflects that the attacker must already hold privileges sufficient to create or configure a custom tool; UI:R reflects that a distinct victim with the mcp_tool_execute capability must deliberately run the attacker's tool. The high confidentiality impact (C:H) comes from the captured token granting access to the victim's data and permitted actions, while the lower integrity and availability impacts (I:L/A:L) reflect that the token's reachable actions are bounded by the victim's own role permissions. The affected CPE cpe:2.3:a:splunk:splunk_mcp_server:*:*:*:*:*:*:*:* is general; EUVD narrows the vulnerable range to the 1.2 line below 1.2.1.
Know when Splunk 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
AV:N/AC:H/PR:H/UI:R/S:U/C:H/I:L/A:L
Attacker needs tool-configuration privileges (PR:H) and a distinct privileged victim must run the tool (UI:R, AC:H); token theft yields full impersonation (C:H) with lesser integrity/availability impact.
Primary rating from Vendor (cisco).
CVSS VectorVendor: cisco
Lifecycle Timeline
1DescriptionCVE.org
In Splunk MCP Server versions below 1.2.1, Splunk MCP Server could send the Splunk platform authentication token of a user who runs a custom Application Programming Interface (API) tool to the URL configured for that tool. If another user controls that URL, they could capture the token and use it to access data and perform actions as the user who ran the tool. Successful exploitation requires a user who holds a role that contains the mcp_tool_execute capability to run a custom API tool configured by another user. For more information see Configure the Splunk MCP Server (https://help.splunk.com/en/splunk-enterprise/mcp-server-for-splunk-platform/1.2/configure-the-splunk-mcp-server) and Managing custom tools in Splunk MCP Server (https://help.splunk.com/en/splunk-enterprise/mcp-server-for-splunk-platform/1.2/managing-custom-tools-in-splunk-mcp-server) in the Splunk documentation.
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.
The (1) TLS and (2) DTLS implementations in OpenSSL 1.0.1 before 1.0.1g do not properly handle Heartbeat Extension packe
Unauthenticated arbitrary file write in Splunk Enterprise (below 10.2.4 and 10.0.7) and Splunk Cloud Platform (below 10.
curl 7.75.0 through 7.76.1 suffers from a use-after-free vulnerability resulting in already freed memory being used when
curl < 7.84.0 supports "chained" HTTP compression algorithms, meaning that a serverresponse can be compressed multiple t
In Splunk Enterprise versions below 9.2.2, 9.1.5, and 9.0.10, a low-privileged user that does not hold the admin or powe
In Splunk Enterprise versions below 9.0.7 and 9.1.2, Splunk Enterprise does not safely sanitize extensible stylesheet la
In versions of Splunk Enterprise below 9.0.5, 8.2.11, and 8.1.14, and Splunk Cloud Platform below version 9.0.2303.100,
In Splunk Enterprise versions below 8.2.9, 8.1.12, and 9.0.2, an authenticated user can execute arbitrary code through t
When curl < 7.84.0 saves cookies, alt-svc and hsts data to local files, it makes the operation atomic by finalizing the
SQLite 1.0.12 through 3.39.x before 3.39.2 sometimes allows an array-bounds overflow if billions of bytes are used in a
A vulnerability exists in curl <7.87.0 HSTS check that could be bypassed to trick it to keep using HTTP. Rated high seve
When doing HTTP(S) transfers, libcurl might erroneously use the read callback (`CURLOPT_READFUNCTION`) to ask for data t
EUVD-2026-94587
GHSA-mgwr-9j5x-6w4x
NVD
CVE.org
EPSS
Vulnrichment
Exploit-DB