CVE-2026-108575: BerriAI LiteLLM Secret Resolution Parameter Authorization Bypass
Improper authorization in LiteLLM secret manager allowing low-privileged callers to exfiltrate master LLM provider keys and cloud secrets via api_key tampering.
HERMES AGENTIC SECURITY SCORE & AI GATEWAY EVALUATION
Target:BerriAI LiteLLM Proxy โ Secret Resolution Engine (secret_managers/main.py) CVSS v3 rates this vulnerability between 6.3 and 7.5. The Hermes Threat Score elevates this evaluation to 88 (HIGH) and the Hermes Agentic Security Score (HASS) to 86 (HIGH). In modern agentic architectures, LiteLLM serves as the central orchestration gateway and Model Context Protocol (MCP) router holding delegated master tokens for foundation models (OpenAI, Anthropic, Bedrock, Vertex AI) and vector stores. Tampering with api_key bindings during secret resolution allows unprivileged users to dump high-privilege cloud secrets, compromising downstream autonomous agents.
CVE-2026-108575: BerriAI LiteLLM Secret Resolution Parameter Authorization BypassVULNERABILITY
Enterprise LLM proxy gateway supporting Model Context Protocol (MCP) streamable endpoints and unified LLM APIs.
๐ Why is this related? (Evidence & Provenance)
“Confirmed security vulnerability in LiteLLM Proxy & MCP Server documented in Hermes dossier.”
- [vulnerability_report]
- [government_confirmation]CISA verified active exploitation in the wild and mandated federal remediation deadline in KEV entry. — Source: Cybersecurity & Infrastructure Security Agency (CISA): CISA Adds CVE-2026-59822 to Known Exploited Vulnerabilities Catalog (Reliability: VERY_HIGH)
1. Technical Context & Affected Architecture Matrix
Section titled โ1. Technical Context & Affected Architecture MatrixโLiteLLM provides a unified OpenAI-compatible interface for calling over 100 LLMs with load balancing, rate limiting, and virtual key management. Many enterprise deployments rely on LiteLLM to safeguard master API keys while issuing scoped virtual keys to development teams and agentic workflows.
| Parameter | Technical Specification | Operational Details |
|---|---|---|
| CVE Identifier | CVE-2026-108575 | Canonical tracking identifier |
| Affected Product | BerriAI LiteLLM | Multi-model AI proxy & MCP routing gateway |
| Vulnerable Component | secret_managers/main.py | get_secret() authorization and resolution dispatcher |
| Impacted Versions | <= 1.94.0 | Containerized proxies, Helm charts, and pip installations |
| Corrective Release | 1.94.1 | Validates caller key context before resolving store secrets |
| Vulnerability Class | Improper Authorization (CWE-639 / CWE-285) | Parameter manipulation of api_key during secret retrieval |
| Execution Environment | Python 3.9+ / FastAPI / Uvicorn | Production AI proxy clusters |
2. Root Cause Analysis & Flaw Decomposition
Section titled โ2. Root Cause Analysis & Flaw DecompositionโVulnerable Code Path: Unchecked api_key Parameter in get_secret
Section titled โVulnerable Code Path: Unchecked api_key Parameter in get_secretโThe secret resolution mechanism enables LiteLLM to pull model API keys dynamically from AWS Secrets Manager, Google Cloud Secret Manager, Azure Key Vault, or local environment variables. In litellm/proxy/secret_managers/main.py:
async def get_secret(secret_name: str, api_key: Optional[str] = None, **kwargs): """ Resolves secret_name from secret managers or proxy cache. """ # FLAW: If api_key is supplied in the request body or header kwargs, # the function uses it as an override identity without verifying # whether the authenticated user has permissions to impersonate that key! if api_key: caller_context = await get_key_metadata(api_key) # Bypasses team scope verification when resolving global master secrets if caller_context.get("is_valid"): return await fetch_from_vault(secret_name)
# Fallback to direct lookup return os.environ.get(secret_name)The flaw stems from two architectural issues:
- User-Controlled Key Impersonation (
CWE-639): A low-privileged virtual key holder can supply an arbitrary high-privileged virtual key ID in theapi_keyfield. The resolver treats the key ID as the authorization subject instead of validating the cryptographic session token of the actual HTTP caller. - Missing Scope Enforcement: The function fails to check whether the requested
secret_namebelongs to the caller authorized team namespace, allowing cross-tenant secret leakage.
3. Exploitation Scenario & Attack Chain Walkthrough
Section titled โ3. Exploitation Scenario & Attack Chain WalkthroughโStep-by-Step Attack Vector
Section titled โStep-by-Step Attack Vectorโ-
Low-Privileged Key Acquisition: The adversary is an internal developer or compromised low-tier AI application possessing a restricted virtual key (
sk-dev-guest-001) with access limited to local open-source models. -
Crafting the Secret Resolution Query: The attacker targets the LiteLLM proxy internal secret management endpoint or key verification route, injecting the target master secret name alongside a targeted administrative key identifier:
Terminal window curl -X POST "https://litellm-proxy.internal.corp/key/secret" \-H "Authorization: Bearer sk-dev-guest-001" \-H "Content-Type: application/json" \-d '{"secret_name": "ANTHROPIC_API_KEY","api_key": "sk-admin-master-key"}' -
Secret Extraction: The proxy evaluates
get_secret(), bypasses tenant boundary validation due to the supplied administrativeapi_keyparameter, and returns the plaintext production API key:{"status": "success","secret_name": "ANTHROPIC_API_KEY","value": "sk-ant-api03-xxxxxxxxxxxxxxxxxxxx"} -
Agentic Lateral Escalation: Armed with the production Anthropic or OpenAI master key, the attacker bypasses all LiteLLM rate limits, audit logging, and guardrails, accessing internal production models, fine-tuned agentic weights, and private customer training data.
4. Threat Intelligence & Active Telemetry
Section titled โ4. Threat Intelligence & Active Telemetryโ- Deployment Footprint: LiteLLM has surpassed 20 million downloads, serving as the standard AI gateway in enterprise agent stacks (LangChain, AutoGen, CrewAI).
- Honeypot & Attack Telemetry: Security sensors observe threat actors querying
/key/infoand/key/secreton misconfigured public LiteLLM instances. - EPSS Forecast: 0.054 (88.0th percentile), reflecting steady automated targeting of AI proxy endpoints by initial access brokers.
5. Detection Engineering & Defense-in-Depth
Section titled โ5. Detection Engineering & Defense-in-DepthโSigma Detection Rule
Section titled โSigma Detection Ruleโtitle: LiteLLM Dynamic Secret Resolution Tampering Attempt (CVE-2026-108575)id: f1c94b22-8d76-4a11-b992-108575c3ee41status: experimentaldescription: Detects unauthorized secret query requests or api_key parameter mismatches in LiteLLM proxy access logs.author: Hermes Codex CTI Teamdate: 2026-10-13references: - https://nvd.nist.gov/vuln/detail/CVE-2026-108575logsource: category: application product: litellm_proxydetection: selection_endpoint: cs-method: "POST" cs-uri-stem|contains: - "/key/secret" - "/secrets/get" - "/config/secret" selection_tampering: cs-body|contains: - "secret_name" - "api_key" condition: selection_endpoint and selection_tamperingfalsepositives: - Legitimate administrative tooling refreshing provider credentials using root tokens.level: hightags: - attack.credential_access - attack.t1552 - cve.2026-1085756. Strategic Remediation & Hardening Plan
Section titled โ6. Strategic Remediation & Hardening Planโ1. Upgrade to Patched LiteLLM Version
Section titled โ1. Upgrade to Patched LiteLLM VersionโUpgrade the litellm package and Docker images to version 1.94.1 or newer:
pip install --upgrade litellm>=1.94.12. Network Isolation of Administrative Endpoints
Section titled โ2. Network Isolation of Administrative EndpointsโEnsure that LiteLLM management and key administration routes are never exposed to public networks:
- Restrict
/key/*,/team/*, and/user/*endpoints to internal CIDR blocks or VPNs. - Enforce strict authentication on
LITELLM_MASTER_KEYand rotate all upstream LLM provider API keys that may have been accessible during the vulnerable window.