UTC

vllm

131 published advisories; version 0.34.0 or later fixes all but 6 of them.

Registry
PyPI
Advisories
131
Latest fix
0.34.0
Exploited in the wild
None listed
Advisories with no fix yet

6 advisories have no fixed release in any version yet. What can be done meanwhile:

General guidance from ShieldPi, not a vendor fix

  • Command or code injection. Remove or disable the affected server or tool, or run it in a sandbox with no secrets in its environment.GHSA-cj47-qj6g-x7r4
  • Unsafe deserialization. Do not let it load files, models or data from sources you do not trust, and keep its RPC and socket ports reachable only from localhost or hosts you trust.GHSA-5vqr-wprc-cpp7 · GHSA-cj47-qj6g-x7r4 · GHSA-pgr7-mhp5-fgjp

Check your machine

Finds vllm in your agent and MCP configs, installed skills, global npm packages, the npx and uvx caches, uv and pipx tools, Claude Code and Cursor. It runs offline and sends nothing anywhere.

pip install "shieldpi>=0.10.0"
shieldpi agent-config . --installed
Provenance

Source: github.com/vllm-project/vllm, named in the package metadata (not verified)

0 of the last 10 releases of vllm were published with build provenance: a signed record of the repository and commit each was built from, which the registry checks before it accepts the release.

  • 0.34.0
  • 0.33.0
  • 0.32.0
  • 0.31.0
  • 0.30.0
  • 0.29.0
  • 0.28.0
  • 0.27.1
  • 0.27.0
  • 0.26.0

Green: published with provenance. Grey: without.

OpenSSF Scorecard has not scored this repository.

Provenance from deps.dev (CC BY 4.0); Scorecard from OpenSSF (CDLA-Permissive-2.0). As of 2026-11-18.

Advisories
AdvisorySeverityAffectedFixed inPublished
/v1/systemone bypasses the --trust-request-chat-template gateGHSA-xgh8-qwjh-q853Medium>= 0.32.0rc0, < 0.34.00.34.02026-11-18
Thinking-budget sparse state aliases onto a co-batched request during a mixed `SWAP`, forcing the reasoning-end token into another request'sGHSA-fpch-3xrf-7q96Medium>= 0.21.0, < 0.28.00.28.02026-11-09
Cross-user token data leakage in async scheduling + Pipeline ParallelismGHSA-v97p-wpxp-px84Medium>= 0.16.0, < 0.21.00.21.02026-11-09
Response amplification in OpenAI validation error handlerGHSA-mwgq-9582-w2vfMedium>= 0.3.0, < 0.30.00.30.02026-11-07
Missing request-wide media budget allows single-request frontend resource exhaustionGHSA-5r3c-5x4v-8gjhHigh>= 0.6.0, < 0.33.00.33.02026-11-02
Unauthenticated EngineCore crash via out-of-range prompt token id under skip_tokenizer_init (incomplete fix of GHSA-25q3-v2hm-8vpf)GHSA-q9v9-qpjp-fc24High>= 0.28.0, < 0.33.00.33.02026-11-02
vLLM Speculators Config verifier.name_or_path Element Directory Traversal Information Disclosure VulnerabilityGHSA-qrj2-j723-m6w8Medium< 0.33.00.33.02026-11-02
vLLM: DoS via unbounded chat_template Jinja2 rendering exhausts the renderer thread poolCVE-2026-115911Medium>= 0.19.0, < 0.32.00.32.02026-10-28
vLLM: Public multimodal SHM handle forgery reaches worker-side `pickle.loads()` with insecure serialization disabledCVE-2026-115912High< 0.32.00.32.02026-10-28
vLLM: Phi-4 multimodal audio encoder materializes and retains full attention matrices: a single max-length audio OOM-crashes the vLLM engineCVE-2026-115913Medium< 0.32.00.32.02026-10-28
vLLM DeepSeek-OCR image processing vulnerable to denial of serviceCVE-2026-115914Medium>= 0.11.1, < 0.32.00.32.02026-10-28
vLLM: Untrusted EPS inputs can trigger denial of service or Ghostscript RCECVE-2026-115915High< 0.32.00.32.02026-10-28
vLLM: Out-of-vocabulary prompt token id passes input validation and triggers a CUDA device-side assert (engine crash / DoS)CVE-2026-115916Medium< 0.32.00.32.02026-10-28
vLLM: Permissive default CORS origins allow cross-origin requests to the vLLM HTTP serverCVE-2026-115917Low< 0.32.00.32.02026-10-28
vLLM: Responses API bypasses trust_request_chat_template for chat_template_kwargsCVE-2026-115918Medium>= 0.16.0, < 0.32.00.32.02026-10-28
vLLM denial of service via prompt embeds on XD-RoPE models (incomplete fix of CVE-2026-55514)GHSA-5q78-j82c-2vr2Medium>= 0.28.0, < 0.30.00.30.02026-10-27
vLLM OpenCVDynamicOpenPanguVideoBackend: request-level num_frames without a ceiling — unauthenticated memory/CPU exhaustion (OOM-kill) on anGHSA-894j-wg8m-63h8Medium>= 0.30.0, < 0.31.00.31.02026-10-27
Quadratic Cohere2 Vision preprocessing DoS via request-controlled max_patchesGHSA-9g3c-x747-c6wrMedium>= 0.10.1, < 0.31.00.31.02026-10-27
vllm p2p never connected zmq_exhaustionCVE-2026-94624Medium< 0.32.00.32.02026-10-27
vLLM: Request-controlled mm_processor_kwargs.code_revision allows remote code executionCVE-2026-114053High>= 0.7.3, < 0.31.00.31.02026-10-27
sanitize_message() address regex is lowercase-only and does not match PIL's uppercase __repr__, leaking live heap addresses to unauthenticatGHSA-m9fg-w9qh-67pcMedium>= 0.14.0rc1, < 0.31.00.31.02026-10-27
Incomplete CVE-2026-34755 remediation can be bypassed by request-level num_framesGHSA-vxqj-p4gw-9h4cMedium>= 0.19.0, < 0.31.00.31.02026-10-27
MiniMax M3 video preprocessing permits request-driven CPU resource exhaustion through min_pixelsGHSA-w4cg-mhc7-mgw3Medium>= 0.24.0, < 0.31.00.31.02026-10-27
vLLM : DoS via unvalidated mm_processor_kwargs in multimodal image processors (Isaac, InternVL, H2OVL, etc.)CVE-2026-114055High>= 0.7.3, < 0.31.00.31.02026-10-23
vLLM: Forged multimodal UUID + P0/P1 processor-cache eviction drift: unauthenticated remote crash of the whole vLLM engineCVE-2026-114050Medium< 0.31.00.31.02026-10-23
Remote DoS in disaggregated multimodal /inference/v1/generate via malicious kwargs_data / mm_placeholdersGHSA-34qw-q2h2-jp3mMedium>= 0.20.0, < 0.30.00.30.02026-10-09
Frame-count cap bypass in VideoMediaIO.load_base64 for video/jpeg via media_io_kwargs.video.num_framesGHSA-4m6g-h9mv-wqjcHigh>= 0.18.0, < 0.31.00.31.02026-10-09
Rust gRPC output_candidates.token_ids bypasses logprob_token_ids limits and can terminate EngineCoreGHSA-6pgw-f3pv-4h8qMedium>= 0.22.0, < 0.24.00.24.02026-10-09
Remote DoS in vLLM via allowed_token_ids Speculative-Decoding Metadata MismatchGHSA-8c65-hq7q-r7jmHigh>= 0.11.1, < 0.18.00.18.02026-10-09
Nemotron-VL processors disable Pillow decompression-bomb guard globally, enabling unauthenticated DoSGHSA-mm38-5g96-7j6xHigh<= 0.8.50.25.02026-10-09
Multimodal chat audio decoding bypasses the decompression-bomb duration limitGHSA-r45p-7qwx-qwjwMedium>= 0.19.0, < 0.24.00.24.02026-10-09
Cross-tenant DoS via uninitialized LogprobsTensors with prompt_logprobsGHSA-rhcx-5729-88vgHigh>= v0.8.5, < 0.31.00.31.02026-10-09
Rust gRPC CandidateTokens.top_n bypasses max_logprobs and can terminate EngineCoreGHSA-rwfw-xvg3-3937Medium>= 0.22.0, < 0.24.00.24.02026-10-09
SSRF via X-Request-Id header drives P2pNcclConnector outbound ZMQ+NCCL pair, leaking KV-cache tensorsGHSA-vfp2-c8pq-v6h6High>= 0.9.2, < 0.24.00.24.02026-10-09
Rust gRPC Generate accepts out-of-vocabulary prompt token IDs and can terminate EngineCoreGHSA-w2v2-5662-8v5qMedium>= 0.22.0, < 0.24.00.24.02026-10-09
Unvalidated stop_token_ids causes engine crash (DoS) and out-of-bounds GPU writeGHSA-w9r6-29q4-pvprMedium>= 0.20.1, < 0.29.00.29.02026-10-09
An oversized min_tokens in one request hangs the vLLM engine while /health stays greenGHSA-4xqp-c3mv-qff7Medium< 0.31.00.31.02026-10-06
V1 InputBatch.condense leaves a stale allowed_token_ids mask on recycled batch rowsGHSA-6cxc-2vcg-w5qcLow< 0.31.00.31.02026-10-06
MOSS-Audio request-controlled processor cache permits remote memory exhaustionCVE-2026-114052High> 0.24.0, < 0.31.00.31.02026-10-06
vLLM: Harmony tool continuations drop `cache_salt` — restoring a cross-tenant prefix-cache membership oracleCVE-2026-105752Low< 0.30.00.30.02026-10-06
vLLM: Mirrored multimodal IPC caches desync after a rejected request — a later request reusing the same media hash trips a receiver assertioCVE-2026-105753Medium< 0.28.00.28.02026-10-06
Caller-controllable values become or collide with shared cache keys, enabling cross-request cache aliasing (2 sites: multimodal EXIF hash anGHSA-rh6f-3x46-j33qMedium< 0.31.00.31.02026-10-06
vLLM: Loose `cache_salt` validation lets a single request kill EngineCore on LMCache-MP deployments — uncaught downstream `ValueError` deniaCVE-2026-105756Medium< 0.30.00.30.02026-10-05
vLLM: Flash late-interaction scoring caches query embeddings under a caller-controlled request id — cross-request integrity break and induceCVE-2026-105755Medium< 0.30.00.30.02026-10-05
vLLM: GLMGA video sampling permits request-driven CPU and memory exhaustionCVE-2026-105760Medium>= 0.23.0rc2, < 0.30.00.30.02026-10-05
vLLM: Structured-output request errors escape the request boundary and terminate the shared EngineCore — engine-fatal denial of service (3 sCVE-2026-105757Medium< 0.30.00.30.02026-10-05
vLLM: Scale-out disaggregated multimodal transport trusts caller-supplied featuresCVE-2026-105754Medium< 0.30.00.30.02026-10-05
vLLM: Qwen2-VL / Qwen3-VL video samplers bound on request-controlled max_frames, which the num_frames ceiling does not reachCVE-2026-105758Medium>= 0.24.0, < 0.30.00.30.02026-10-05
Crafted IAMF audio upload reaches a PyAV/FFmpeg native heap overflow through the speech transcription path — denial of serviceGHSA-m52c-39rh-f3gpMedium< 0.29.00.29.02026-09-28
vLLM: Request-selected PyNvVideoCodec GPU decode bypasses static VRAM reservationCVE-2026-69147Medium< 0.28.00.28.02026-09-17
vLLM: Unauthenticated audio decompression-bomb DoS in /v1/chat/completionsCVE-2026-57173Medium<= 0.23.00.24.02026-09-16
Disaggregated generate skips decoder prompt-length validation for some multimodal processorsGHSA-3mqx-f33v-vgp9Medium< 0.29.00.29.02026-09-12
Incomplete artifact pin propagation in FunAudioChat and Tarsier2GHSA-hhv2-872h-628qMedium>= 0.22.1, < 0.28.00.28.02026-09-12
Sampler Subclass Counter Shadowing Bypasses PyNvVideoCodec Decoder Limits and GPU Memory AccountingCVE-2026-100649Low<0.29.00.29.02026-09-12
Uncontrolled resource consumption: multimodal chat audio decoding ignores `VLLM_MAX_AUDIO_CLIP_FILESIZE_MB`CVE-2026-100648Medium< 0.29.00.29.02026-09-12
Remote media is fetched and fully materialized before the documented media size / item limits are enforced — remote pre-inference memory andGHSA-p6g9-7v3x-m8mvMedium<= 0.29.0No fix yet2026-09-12
Rust HTTP/gRPC stop_token_ids bypass Python vocab-bound fix and can terminate EngineCoreGHSA-qff2-492f-9fm4Medium>= 0.22.0, < 0.24.00.24.02026-09-12
Out-of-range `stop_token_ids` with `min_tokens` can kill vLLM EngineCoreGHSA-v5gm-qgmv-gc6cMedium< 0.29.00.29.02026-09-12
vLLM DoS via unbounded `cache_salt` length: multi-hundred-MB salt stalls the single EngineCore scheduler thread (CPU exhaustion)CVE-2026-100647Medium<0.29.00.29.02026-09-12
vLLM: SSRF + arbitrary local file read in MiMoV2OmniMultiModalProcessor `_fetch_image` and audio loader bypass MediaConnector protectionsCVE-2026-73560Medium< 0.26.00.26.02026-09-08
vLLM: Cross-User Data Leak VulnerabilityCVE-2026-73558Medium< 0.27.00.27.02026-09-08
vLLM: ReDoS via structured_outputs.regex in the lm-format-enforcer backend (no compile timeout) — missed sibling of GHSA-rwxx-mrjm-wc2mCVE-2026-73556Medium< 0.26.00.26.02026-09-04
vLLM: Derender endpoints decode caller-supplied GenerateResponse token IDs without output boundsCVE-2026-71486Medium< 0.26.00.26.02026-09-04
vLLM: Unauthenticated Internal Path and Username Disclosure via Validation Error MessagesCVE-2026-73555Medium< 0.26.00.26.02026-09-04
vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by concurrent prompt partsCVE-2026-73557Medium>= 0.21.0, < 0.26.00.26.02026-09-04
Denial of service: a negative token id in `/v1/embeddings` or `/pooling` input kills the vLLM engine via a CUDA device-side assertionCVE-2026-93592High< 0.28.00.28.02026-09-03
LlavaOnevision2 processor loader executes attacker model code with `trust_remote_code=False` (inert `trust_remote_code` kwarg to `transformeCVE-2026-90553High< 0.28.00.28.02026-08-28
Denial of Service NanoNemoTronVL Video Audio Extraction BombCVE-2026-90554Medium>=0.10.2, < 0.28.00.28.02026-08-28
Speech-to-text audio decode duration limit bypass via forged header sample rateCVE-2026-90555Medium< 0.28.00.28.02026-08-28
vLLM: Completion prompt lists fan out into unbounded engine requestsCVE-2026-73559Medium>= 0.19.0, < 0.26.00.26.02026-08-13
vLLM Unauthenticated Requests Exploit DeepStream Backend Confusion for DoSCVE-2026-78684Medium< 0.27.00.27.02026-08-11
vLLM denial of service via prompt embeds on M-RoPE modelsCVE-2026-55514High>= 0.12.0, < 0.24.00.24.02026-07-20
vLLM: Processing differential in multi-channel audio downmixing enables hidden-input/moderation bypass for audio modelsCVE-2026-34760Medium>= 0.5.5, < 0.18.00.18.02026-07-17
vLLM has Remote DoS via Invalid Recovered Token ReinjectionCVE-2026-54234High>= 0.17.1, < 0.24.00.24.02026-07-17
vLLM: ReDoS via structured_outputs.regex compiled without timeout in xgrammar and outlines backendsCVE-2026-55574High< 0.24.00.24.02026-07-17
vLLM: Speech-to-text upload size limit is enforced after full UploadFile readCVE-2026-55646Medium>= 0.22.0, < 0.24.00.24.02026-07-17
vLLM: GGUF dequantize kernel int truncation exposes uninitialized GPU memory in multi-tenant servingCVE-2026-53923Medium>= 0.5.5, < 0.24.00.24.02026-06-17
vLLM: OOM Denial of Service via Audio Decompression BombCVE-2026-54233Medium<= 0.23.00.24.02026-06-17
vLLM: temperature=NaN and temperature=Infinity bypass validation and propagate to GPU kernelsCVE-2026-54235Medium>= 0.8.5, <= 0.23.00.24.02026-06-17
vLLM: image EXIF Rotation & PNG tRNS Transparency Not Normalized, Causing Mismatch Between Model Input and ExpectationsCVE-2026-12491Medium>= 0.11.0, < 0.24.00.24.02026-06-17
vLLM: incomplete CVE-2026-22778 fix leaks PIL repr addresses via Anthropic routerCVE-2026-54236Medium<= 0.23.00.24.02026-06-17
vLLM: OpenAI auth bypassCVE-2026-48746Critical>= 0.3.0, < 0.22.00.22.02026-06-16
vLLM: Security Check Bypass via assert Statement in Activation Function Loading Allows Arbitrary Code ExecutionCVE-2026-41523High< 0.22.00.22.02026-06-16
vLLM is vulnerable to an Out-of-Memory (OOM) Denial of Service (DoS) attack due to unbounded frame count processing in the `VideoMediaIO.loaCVE-2026-5497High>= 0.8.0, < 0.19.00.19.02026-06-11
vLLM's Artifact Pin Decay allows pinned deployments to load unpinned code, weights, and processorsCVE-2026-47155Medium< 0.22.00.22.02026-06-10
vllm has Improper Resource Shutdown or ReleaseCVE-2026-9540Medium<= 0.19.0No fix yet2026-05-26
vLLM: extract_hidden_states speculative decoding crashes server on any request with penalty parametersCVE-2026-44223Medium>= 0.18.0, < 0.20.00.20.02026-05-06
vLLM Vulnerable to Remote DoS via Special-Token PlaceholdersCVE-2026-44222Medium>= 0.6.1, < 0.20.00.20.02026-05-05
vLLM makes Use of Uninitialized ResourceCVE-2026-7141Low< 0.19.10.19.12026-04-27
vLLM: Unauthenticated OOM Denial of Service via Unbounded `n` Parameter in OpenAI API ServerCVE-2026-34756Medium>= 0.1.0, < 0.19.00.19.02026-04-03
vLLM: Server-Side Request Forgery (SSRF) in `download_bytes_from_url `CVE-2026-34753Medium>= 0.16.0, < 0.19.00.19.02026-04-03
vLLM: Denial of Service via Unbounded Frame Count in video/jpeg Base64 ProcessingCVE-2026-34755Medium>= 0.7.0, < 0.19.00.19.02026-04-03
vLLM has Hardcoded Trust Override in Model Files Enables RCE Despite Explicit User Opt-OutCVE-2026-27893High>= 0.10.1, < 0.18.00.18.02026-03-27
vLLM has SSRF Protection BypassCVE-2026-25960Medium>= 0.15.1, < 0.17.00.17.02026-03-09
vLLM has RCE In Video ProcessingCVE-2026-22778Critical>= 0.8.3, < 0.14.10.14.12026-02-02
vLLM vulnerable to Server-Side Request Forgery (SSRF) through MediaConnectorCVE-2026-24779High< 0.14.10.14.12026-01-28
vLLM affected by RCE via auto_map dynamic module loading during model initializationCVE-2026-22807High>= 0.10.1, < 0.14.00.14.02026-01-21
vLLM is vulnerable to DoS in Idefics3 vision models via image payload with ambiguous dimensionsCVE-2026-22773Medium>= 0.6.4, < 0.12.00.12.02026-01-13
vLLM introduced enhanced protection for CVE-2025-62164CVE-2026-56340High>= 0.10.2, < 0.13.00.13.02026-01-08
vLLM vulnerable to remote code execution via transformers_utils/get_configCVE-2025-66448High< 0.11.10.11.12025-12-02
vLLM vulnerable to DoS via large Chat Completion or Tokenization requests with specially crafted `chat_template_kwargs`CVE-2025-62426Medium>= 0.5.5, < 0.11.10.11.12025-11-20
vLLM deserialization vulnerability leading to DoS and potential RCECVE-2025-62164High>= 0.10.2, < 0.11.10.11.12025-11-20
vLLM vulnerable to DoS with incorrect shape of multimodal embedding inputsCVE-2025-62372High>= 0.5.5, < 0.11.10.11.12025-11-20
vLLM is vulnerable to Server-Side Request Forgery (SSRF) through `MediaConnector` classCVE-2025-6242High>= 0.5.0, < 0.11.00.11.02025-10-07
vLLM: Resource-Exhaustion (DoS) through Malicious Jinja Template in OpenAI-Compatible ServerCVE-2025-61620Medium>= 0.5.1, < 0.11.00.11.02025-10-07
vLLM is vulnerable to timing attack at bearer authCVE-2025-59425High< 0.11.00.11.02025-10-07
vLLM has remote code execution vulnerability in the tool call parser for Qwen3-CoderCVE-2025-9141High>= 0.10.0, < 0.10.1.10.10.1.12025-08-21
vllm API endpoints vulnerable to Denial of Service AttacksCVE-2025-48956High>= 0.1.0, < 0.10.1.10.10.1.12025-08-21
Potential Timing Side-Channel Vulnerability in vLLM’s Chunk-Based Prefix CachingCVE-2025-46570Low< 0.9.00.9.02025-05-28
vLLM DOS: Remotely kill vllm over http with invalid JSON schemaCVE-2025-48942Medium>= 0.8.0, < 0.9.00.9.02025-05-28
vLLM allows clients to crash the openai server with invalid regexCVE-2025-48943Medium>= 0.8.0, < 0.9.00.9.02025-05-28
vLLM has a Weakness in MultiModalHasher Image Hashing ImplementationCVE-2025-46722Medium>= 0.7.0, < 0.9.00.9.02025-05-28
vLLM vulnerable to Regular Expression Denial of ServiceGHSA-j828-28rj-hfhpMedium>= 0.6.3, < 0.9.00.9.02025-05-28
vLLM Tool Schema allows DoS via Malformed pattern and type FieldsCVE-2025-48944Medium>= 0.8.0, < 0.9.00.9.02025-05-28
vLLM has a Regular Expression Denial of Service (ReDoS, Exponential Complexity) Vulnerability in `pythonic_tool_parser.py`CVE-2025-48887Medium>= 0.6.4, < 0.9.00.9.02025-05-28
vLLM Allows Remote Code Execution via PyNcclPipe Communication ServiceCVE-2025-47277Critical>= 0.6.5, < 0.8.50.8.52025-05-20
Remote Code Execution Vulnerability in vLLM Multi-Node Cluster ConfigurationCVE-2025-30165High>= 0.5.2, < 0.10.00.10.02025-05-06
Data exposure via ZeroMQ on multi-node vLLM deploymentCVE-2025-30202High>= 0.5.2, < 0.8.50.8.52025-04-29
vLLM Vulnerable to Remote Code Execution via Mooncake IntegrationCVE-2025-32444Critical>= 0.6.5, < 0.8.50.8.52025-04-29
vLLM: Quadratic Time Complexity in Input Token Processing​ leads to denial of serviceCVE-2025-46560Medium>= 0.8.0, < 0.8.50.8.52025-04-29
CVE-2025-24357 Malicious model remote code execution fix bypass with PyTorch < 2.6.0GHSA-ggpf-24jw-3fcwCritical< 0.8.00.8.02025-04-23
vLLM vulnerable to Denial of Service by abusing xgrammar cacheGHSA-hf3c-wxg2-49q9Medium>= 0.6.5, < 0.8.40.8.42025-04-15
vLLM Deserialization of Untrusted Data vulnerabilityCVE-2024-11041Critical<= 0.6.2No fix yet2025-03-20
vLLM allows Remote Code Execution by Pickle Deserialization via AsyncEngineRPCServer() RPC server entrypointsCVE-2024-9053Critical<= 0.6.0No fix yet2025-03-20
vLLM deserialization vulnerability in vllm.distributed.GroupCoordinator.recv_objectCVE-2024-9052Critical<= 0.8.1No fix yet2025-03-20
vLLM denial of service via outlines unbounded cache on diskCVE-2025-29770Medium< 0.8.00.8.02025-03-19
vLLM Allows Remote Code Execution via Mooncake IntegrationCVE-2025-29783Critical>= 0.6.5, < 0.8.00.8.02025-03-19
vLLM uses Python 3.12 built-in hash() which leads to predictable hash collisions in prefix cacheCVE-2025-25183Low< 0.7.20.7.22025-02-06
vllm: Malicious model to RCE by torch.load in hf_model_weights_iteratorCVE-2025-24357High< 0.7.00.7.02025-01-27
vLLM denial of service vulnerabilityCVE-2024-8768High< 0.5.50.5.52024-09-17
vLLM Denial of Service via the best_of parameterCVE-2024-8939Medium<= 0.5.0.post1No fix yet2024-09-17

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].