CVE-2026-115886

validatedZero produced an exploit or a check.

GHSA-2wqp-q3c6-ccj5 ↗EUVD-2026-109184 ↗

check availableZero validated a working check for it.not proven exploitableZero has a detection check, but hasn't produced a full working exploit yet.
cve.orgpublic pocExploit code has been published by someone. Source signals can also carry a PoC (Proof of Concept).+5.7dkevAttacks confirmed in the real world.+6d
MLflow >= 3.12.0rc0, < 3.17.0
fixed in 3.17.0
ETK-2026-2980Zero's internal tracking ID, independent of the CVE number., permanent identifier
pkg:pypi/mlflow
+5.7dTTV
<1mAI analysisAnalysis: how long the AI spent reversing this signal.
Check if you are vulnerable
timeline
0 = trigger signal
cve.org publication−5.1dtrigger signal +0dcheck available+0.6dpublic pocagent stoppedkev+1.0d
−4d
−2d
0

description

MLflow versions from 3.12.0rc0 through releases before 3.17.0 can allow scorer-related endpoints to trigger server-originated requests to attacker-selected destinations. Exploitation requires an unauthenticated MLflow server with the optional trulens and trulens-providers-litellm packages installed.

mitigation

Upgrade MLflow to version 3.17.0 or later. Until upgraded, require authentication for MLflow endpoints, remove the optional affected dependencies where not needed, and restrict outbound network access from the tracking server.

fixed in 3.17.0

check
available
exploit
no
automationDeterministic: a check that runs on its own, the same request on every target. Contextual: an agentic technique that needs to be applied per target, for bugs that depend on contextual flows.
deterministic
auth requiredWhether an attacker needs an account first, based on the CVSS vector's Privileges Required.
no
EPSSEstimate of the chance this CVE is exploited in the next 30 days. Refreshed daily.
–
agent run time
<1m
cve.org record, t = 0The moment cve.org publishes the CVE in full.
2026-10-23 21:17Z
Δ to public pocExploit code has been published by someone. Source signals can also carry a PoC (Proof of Concept).
+5.7d
Δ to kevAttacks confirmed in the real world.
+6d

outcome

Upstream: a detection for this already ships elsewhere, so we didn't generate a new one. The credit goes to whoever did.

source signals

Public signals that triggered or updated the pipeline. These are inputs for the deltas above.

  1. 2026-10-23 21:17Z
  2. 2026-10-29
    vulncheck_kev: trigger signal, started the run
  3. 2026-10-29 00:00Z
  4. 2026-10-29 15:04Z

this entry in the feed

GET /api/v1/feed/CVE-2026-115886

{
  "etk_id": "ETK-2026-2980",
  "cve_id": "CVE-2026-115886",
  "display_id": "CVE-2026-115886",
  "ghsa_id": "GHSA-2wqp-q3c6-ccj5",
  "euvd_id": "EUVD-2026-109184",
  "status": "published",
  "anchor_state": "anchored",
  "title": "Server-side request forgery in MLflow scorer endpoints",
  "summary": "MLflow versions from 3.12.0rc0 through releases before 3.17.0 can allow scorer-related endpoints to trigger server-originated requests to attacker-selected destinations. Exploitation requires an unauthenticated MLflow server with the optional trulens and trulens-providers-litellm packages installed.",
  "mitigation": "Upgrade MLflow to version 3.17.0 or later. Until upgraded, require authentication for MLflow endpoints, remove the optional affected dependencies where not needed, and restrict outbound network access from the tracking server.",
  "vendor": "MLflow",
  "product": "mlflow",
  "affected": [
    {
      "affected_notes": [],
      "affected_ranges": [
        ">= 3.12.0rc0, < 3.17.0"
      ],
      "ecosystem": "pip",
      "fixed_versions": [
        "3.17.0"
      ],
      "package_name": "mlflow",
      "purl": "pkg:pypi/mlflow"
    }
  ],
  "cwes": null,
  "cvss_score": 6.5,
  "cvss_src": "other",
  "cvss_vector": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N",
  "epss_score": null,
  "epss_percentile": null,
  "epss_asof": null,
  "automation": "deterministic",
  "check": true,
  "check_confidence_basis": "upstream",
  "exploit": false,
  "pipeline_minutes": 0.0027078,
  "t0_at": "2026-10-23T21:17:21.462000+00:00",
  "signals": {
    "check": {
      "at": "2026-10-29T15:04:01.790983+00:00",
      "days": 5.7407445484143516,
      "is_lower_bound": false,
      "measured_at": "2026-10-23T21:17:21.462000+00:00",
      "url": null
    },
    "kev": {
      "at": "2026-10-29T00:00:00+00:00",
      "days": 6,
      "is_lower_bound": false,
      "measured_at": "2026-10-23T21:17:21.462000+00:00",
      "precision": "day",
      "url": "https://vulncheck.com/cve/CVE-2026-115886"
    },
    "public_poc": {
      "at": "2026-10-29T15:04:01.803511+00:00",
      "days": 5.740744693414352,
      "is_lower_bound": false,
      "measured_at": "2026-10-23T21:17:21.462000+00:00",
      "url": "https://github.com/projectdiscovery/nuclei-templates/blob/main/http/cves/2026/CVE-2026-115886.yaml"
    },
    "trigger_signal": {
      "at": "2026-10-29T00:00:00+00:00",
      "days": 5.1129460416666666,
      "is_lower_bound": false,
      "measured_at": "2026-10-23T21:17:21.462000+00:00",
      "url": "https://vulncheck.com/cve/CVE-2026-115886"
    }
  }
}

published 2026-10-29 15:03Z
last change 2026-10-29 16:58Z

Need the technical details behind an entry?

Every entry here shows what we attempted and when. The technical details, the generated exploit that fired it and the evidence behind the verdict are not public. Tell us about your use case and we’ll get back to you about extended API access.