Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata – not a confirmed exploit chain
-
AccessReach Floci API Gateway endpoint
-
DeliveryCreate REST API with MOCK integration
-
ExploitCraft VTL template using $util reflection
-
ExecutionInvoke API to render template
-
PersistReflect to Runtime/ProcessBuilder
-
ImpactExecute OS commands in Floci JVM
Vulnerability Assessment
| Weakness |
CWE
OWASP Top 10:2025
A05:2025 - Injection
ASVS 5.0
V1 - Encoding and Sanitization Code injection prevention, sandbox exec
MITRE ATT&CK
2 techniques across 2 tactics: Execution, Initial 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 pre-2.2.0 Floci instance whose API Gateway service (or the equivalent AppSync resolver path) is network-reachable to the attacker, and whose VTL template processing runs on the pre-patch engine with no reflection sandbox. The concrete prerequisite is the ability to create/define a REST API with a MOCK integration carrying a Velocity mapping template that uses $util reflection to reach Runtime or ProcessBuilder - a capability the vulnerable engine exposes to unauthenticated callers per the CVSS PR:N vector. Limiting factors: the target must be running an affected version (1.1.0 up to but not including 2.2.0); after the 2.2.0 fix, templates render inside SecureUberspector which blocks Class, ClassLoader, Runtime, ProcessBuilder, System, Thread, and java.io.File, plus loop/output/time limits, neutralizing the reflection path. Because Floci is primarily a local/CI developer emulator, realistic exposure depends on whether its control plane is bound beyond localhost. |
| Impact | All exploitability signals align toward high real-world risk. The vendor-supplied CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:N/UI:N) describes network-reachable, low-complexity, unauthenticated exploitation with high confidentiality, integrity, and availability impact (VC:H/VI:H/VA:H), scoring 9.3, and the description corroborates this - unauthenticated OS command execution in the Floci JVM via a crafted Velocity (VTL) mapping template. Publicly available exploit code exists (ghostx-86 PoC) and a vendor patch (2.2.0) is available, and the finding was reported by VulnCheck, a credible source. However, the input provides NO EPSS score and NO CISA KEV listing, so this is 'publicly available exploit code exists' - NOT confirmed active exploitation; do not treat it as in-the-wild. The dominant counterweight is deployment context: Floci is a niche open-source AWS-service emulator (typically a developer/CI tool) rather than ubiquitous production infrastructure, so despite the near-maximal CVSS, aggregate real-world exposure is constrained by how narrowly the product is deployed and whether its API Gateway/AppSync control plane is network-reachable. Net: a genuine high-severity RCE that should be patched promptly where Floci is exposed, but likely targeted/limited rather than mass-exploited given the product footprint. |
| Exploit ScenarioDerived | An attacker with network access to a pre-2.2.0 Floci instance's API Gateway control plane creates a REST API with a MOCK integration whose VTL mapping template uses $util-based reflection to reach java.lang.Runtime or ProcessBuilder. Invoking the API renders the template and executes attacker-chosen OS commands inside the Floci JVM, with no authentication required. A public proof-of-concept for this exact API Gateway VTL RCE path is available. |
Get an e-mail when this CVE is exploited, patched or re-scored. Free account.
Alert meAnalysisAIv1 – Oct 10, 2026
Unauthenticated remote code execution affects Floci versions 1.1.0 through releases prior to 2.2.0, where the API Gateway VTL template engine fails to sandbox Velocity template rendering. An attacker who can reach the Floci control plane can define a REST API with a MOCK integration whose mapping template uses $util reflection to reach Runtime or ProcessBuilder, executing arbitrary OS commands inside the Floci JVM - a CVSS 9.3 (CVSS 4.0: AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H) impact confirmed as unauthenticated by the PR:N vector. Publicly available exploit code exists, though no public exploit identified at time of analysis as confirmed actively exploited (CISA KEV), and a vendor-released fix is available in Floci 2.2.0. Because Floci is primarily a local and CI developer emulator of AWS services, practical exposure depends heavily on whether its API Gateway/AppSync endpoints are bound beyond localhost.
RemediationAI
Upgrade to Floci 2.2.0 or later; the release is published at https://github.com/floci-io/floci/releases/tag/2.2.0 and the hardening change is contained in commit 144b90e6fa861a5ac882e99a090d503da01945c4, which introduces the SecureUberspector reflection sandbox plus loop, output-size, and wall-clock limits for both the API Gateway and AppSync VTL paths. If immediate upgrade is not possible, treat the Floci control plane as a privileged service: bind the API Gateway and AppSync listeners to loopback only rather than 0.0.0.0, so remote callers cannot submit REST API definitions or mapping templates - this eliminates the attack vector entirely at the cost of breaking use by remote CI runners or shared test environments, which would instead need a loopback tunnel or an authenticating reverse proxy. As a secondary containment measure, run Floci under a least-privilege service account or in an unprivileged container with a read-only root filesystem and no host mounts, so that any successful Runtime/ProcessBuilder invocation cannot escalate to meaningful host access; note this does not prevent in-JVM command execution or data exfiltration from the emulator's own state. Also verify that the emulator's administrative/control-plane port is not exposed through ingress rules, and monitor for unexpected outbound connections or child processes spawned by the Floci JVM, since the reflection chain produces genuine OS process creation. Review the advisory at https://github.com/floci-io/floci/security/advisories/GHSA-3p4c-wp7w-mgjx for vendor-published guidance and confirm whether your deployment uses the API Gateway or AppSync template surfaces, as both must be considered exposed until 2.2.0 is applied.
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 rooted in CWE-94 (improper control of generation of code / code injection) within Floci's VtlTemplateEngine, the component that renders Velocity Template Language (VTL) mapping templates for the emulated API Gateway and AppSync services. VTL is the same template dialect AWS uses for API Gateway integration mapping templates and AppSync resolvers, and Velocity's expression language permits method invocation through the $util context object. In the vulnerable code path (VtlTemplateEngine.java, roughly lines 74-114 at commit 560b3e4aae61cea3ac7d4d51843bdffd0e3085fd), templates were evaluated without a restrictive Uberspector, so reflection chains from $util could enumerate and invoke classes such as java.lang.Runtime or java.lang.ProcessBuilder and spawn OS processes with the privileges of the Floci JVM. The 2.2.0 fix hardens the same engine by rendering templates inside a reflection-restricted SecureUberspector sandbox that blocks Class, ClassLoader, Runtime, ProcessBuilder, System, Thread, java.io.File and related classes/packages, and additionally bounds execution with FLOCI_SERVICES_APIGATEWAY_VTL_MAX_LOOPS / APPSYNC equivalents (default 10000), FLOCI_SERVICES_*_VTL_MAX_OUTPUT_CHARS (default 1048576, enforced by a new BoundedWriter), and FLOCI_SERVICES_*_VTL_TIMEOUT_MILLIS (default 5000), plus a SandboxedForeach directive that checks a VtlExecutionGuard deadline on each loop iteration. Those controls also mitigate template-driven denial-of-service such as runaway #foreach loops, which are separate from the reflection-based command-execution primitive described here.
Know when Floci 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:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Unauthenticated network template injection (PR:N/AV:N/AC:L) yields OS command execution within the Floci JVM, giving full C/I/A impact; scope kept Unchanged as execution stays within the vulnerable process context.
Primary rating from Vendor (VulnCheck).
CVSS VectorVendor: VulnCheck
DescriptionCVE.org
Floci 1.1.0 before 2.2.0 contains a code injection vulnerability in VtlTemplateEngine that allows unauthenticated attackers to execute commands via unrestricted Velocity mapping templates. Attackers can create a REST API with a MOCK integration whose template uses $util reflection to reach Runtime or ProcessBuilder, executing OS commands in the Floci JVM.
Technical POC Analysis
CWE-94 code injection.
Full technical POC analysis - root cause, step-by-step exploitation and patch breakdown - is part of the Pro and Business plans.
Latest POC
1 public PoC link from github
Sign in to open the public PoC links collected for this CVE.
Executive ViewAI
Floci, the AWS service emulator our engineering and CI pipelines rely on, is affected by a critical unauthenticated remote code execution flaw in versions 1.1.0 through 2.1.x, where a reachable API Gateway endpoint lets an attacker run arbitrary operating system commands on the host. The realistic organizational exposure is not a customer-facing service but developer workstations and build runners, which typically hold cloud credentials, source code, and pipeline secrets, making a compromise a data and supply-chain risk. Publicly available exploit code exists and a vendor-released fix is available in Floci 2.2.0, so urgency depends primarily on whether these endpoints are network-reachable beyond localhost.
Apply vendor patch.
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
Code injection occurs when an application accepts user input and passes it directly into a language interpreter or evaluator without proper sanitization.
Remote Code Execution represents the critical moment when an attacker successfully runs arbitrary code on a target system without physical access.
EUVD-2026-96394
GHSA-c3c4-9g85-pfvc
NVD
CVE.org
EPSS
Exploit-DB