PYSEC-2026-101

An issue in ollama v.0.12.10 allows a remote attacker to cause a denial of service via the GGUF decoder...

PYSEC-2026-116

OpenCTI is an open source platform for managing cyber threat intelligence knowledge and observables. Prior to version 6.8.1, the GraphQL mutation "WorkspacePopoverDeletionMutation" allows users to delete workspace-related objects such as dashboards and investigation cases. However, the mutation...

PYSEC-2025-56

OctoPrint provides a web interface for controlling consumer 3D printers. In versions up to and including 1.10.3, OctoPrint has a vulnerability that allows an attacker to bypass the login redirect and directly access the rendered HTML of certain frontend pages. The primary risk lies in potential...

PYSEC-2025-196

A vulnerability classified as problematic has been found in PyTorch 2.6.0. Affected is the function torch.jit.jitmodulefromflatbuffer. The manipulation leads to memory corruption. Local access is required to approach this attack. The exploit has been disclosed to the public and may be used...

PYSEC-2025-9

A remote code execution vulnerability exists in invoke-ai/invokeai versions 5.3.1 through 5.4.2 via the /api/v2/models/install API. The vulnerability arises from unsafe deserialization of model files using torch.load without proper validation. Attackers can exploit this by embedding malicious cod...

PYSEC-2024-38

FastAPI is a web framework for building APIs with Python 3.8+ based on standard Python type hints. When using form data, python-multipart uses a Regular Expression to parse the HTTP Content-Type header, including options. An attacker could send a custom-made Content-Type option that is very...

PYSEC-2021-112

An integer overflow exists in pywin32 prior to version b301 when adding an access control entry ACE to an access control list ACL that would cause the size to be greater than 65535 bytes. An attacker who successfully exploited this vulnerability could crash the vulnerable process...

Open WebUI's API key endpoint restrictions bypassed via `x-api-key` header — full message processing on restricted endpoints

SummaryOpen WebUI allows admins to restrict which API endpoints an API key can access. When an API key is restricted from /api/v1/messages, requests using the Authorization: Bearer sk-... header are correctly blocked with 403. However, the same key sent via the x-api-key header bypasses the...

Pillow JPEG2000 tiled decode retains a growing scratch buffer and can be used for denial of service

Summarysrc/libImaging/Jpeg2KDecode.c:853 accumulates totalcomponentwidth across every tile in a JPEG2000 image instead of recomputing it per tile. That accumulated value is then used in the tilebytes calculation at src/libImaging/Jpeg2KDecode.c:868, which can make the decoder grow state-buffer vi...

LiteLLM: Local file read via request-supplied OIDC file references

ImpactLiteLLM's /health/testconnection endpoint resolved request-supplied environment and OIDC file references in litellmparams. A proxy administrator, or another privileged caller with permission to test model connections, could cause LiteLLM to read files from the local filesystem via an...

PocketSphinx: Buffer overflows in language and acoustic model loading code

ImpactThe trie language model code introduced in PocketSphinx 5prealpha failed to check various boundary conditions when reading the headers of ARPA, DMP, and binary format language model files. In the case of invalid, corrupted or malicious input files, this could lead to stack and heap buffer...

Prompty: Arbitrary file read via file reference expansion

SummaryPrompty loaders expanded $file:... references in .prompty frontmatter without enforcing that the resolved path stayed within an authorized directory. An attacker-controlled prompt file could use path traversal or an absolute path to cause the host application to read files accessible to th...

vLLM has Remote DoS via Invalid Recovered Token Reinjection

SummaryA frontend-legal multi-request speculative workload can make vLLM produce an out-of-vocabulary recovered token equal to vocabsize, convert that value to -1 when choosing the next live token for a request, and then feed that -1 back into the next drafter input ids. On Qwen3 GPTQ this reache...

MCP Python SDK: Experimental task handlers allow any client to access and cancel other clients' tasks

SummaryIn affected versions, the default request handlers installed by the experimental tasks feature server.experimental.enabletasks did not check which session created a task before acting on it. On a server with more than one connected client, any client could observe, read results from, and...

DIRAC: Pilot code downloaded over unverified HTTPS connection

SummaryThe second stage pilot pilot.tar is downloaded by the initial wrapper script without any verification of the webservers' SSL certificate and the contained script is subsequently executed. The checksum is tested, but the reference checksum file is downloaded over the same unvalidated channe...

garminconnect Has Insecure Permission Assignment for Garmin OAuth Token Store

Insecure Permission Assignment for Garmin OAuth Token Store Summarygarminconnect ≤ 0.3.4 wrote its OAuth token store to disk without restricting file-system permissions. Under the default Linux umask 022 the token file garmintokens.json was created world-readable 0o644. The file contains the DI...

LangBot: Authenticated RCE Via MCP Configuration

SummaryAny authenticated user can achieve arbitrary command execution on the LangBot servers through changing the MCP Server Configuration by added an "STDIO" MCP with an arbitrary command. DetailsThe repository uses StdioServerParameters which is based on Anthropic's modelcontextprotocol open...

PraisonAI SandlockSandbox falls back to unrestricted subprocess execution when Landlock is unavailable

Summarypraisonai.sandbox.SandlockSandbox is documented and implemented as the kernel-enforced sandbox backend for untrusted code. Its SandboxConfig.native path lets callers configure allowed filesystem paths and network=False.On systems where the optional sandlock module imports but reports that...

netlicensing-mcp: REST Path Traversal Bypasses Token Redaction

REST Path Traversal Bypasses Token Redaction in netlicensing-mcp SummaryThe netlicensinggetproduct MCP tool in netlicensing-mcp interpolates a caller-controlled productnumber argument directly into a REST URL path without any validation. Passing ../token as the product number causes httpx to...

PraisonAI: PRAISONAI_CALL_AUTH=disabled environment variable unconditionally disables authentication

SummarySetting PRAISONAICALLAUTH=disabled completely disables all authentication on the /api/v1/agents/id/invoke endpoint. This bypass is advertised in the application's own error messages, making it likely to appear in production Docker and Compose configurations. Detailspython...

praisonai-platform 0.1.4 still boots on the hardcoded JWT secret dev-secret-change-me (default-open production guard)

- Affected: praisonai-platform PyPI CWE-287 Improper Authentication OverviewGHSA-3qg8-5g3r-79v5 Critical reported that praisonai-platform's JWT signing secret defaulted to the hardcoded literal "dev-secret-change-me", and that the production guard meant to prevent this was default-open it only...

PraisonAI: Missing Authentication for Critical Function and Improper Neutralization of Special Elements used in an OS Command ('OS Command Injection') in praisonai

Unauthenticated PraisonAI UI MCP connect endpoint executes attacker-chosen local commands SummaryPraisonAI v4.6.48 exposes the PraisonAIUI MCP client management API through the default UI host apps without authentication. A remote unauthenticated client can send POST /api/mcp/connect with a comma...

PraisonAI: Missing ownership check on DELETE endpoints allows members to delete others' content in Platform API

SummaryA workspace member can permanently delete any resource — projects, agents, issues, labels, issue dependencies, and issue-label attachments — created by the workspace owner or other members. All six content DELETE endpoints enforce workspace membership but perform no ownership or role check...

praisonaiagents: SSRF guard validates literal IPs only and never resolves DNS

praisonaiagents: SSRF guard validates literal IPs only and never resolves DNSResearcher: Kai Aizen — SnailSploit @SnailSploit, Adversarial & Offensive Security ResearchTarget: https://github.com/MervinPraison/PraisonAIWeakness: CWE-918 Server-Side Request Forgery SSRF.--- SummaryThe SSRF guard...

PraisonAI Slack app_mention bypasses configured user/channel authorization

PraisonAI Slack appmention bypasses configured user/channel authorization SummaryPraisonAI's Slack bot applies its configured allowedusers,allowedchannels, and unknown-user pairing policy in the normal Slackmessage event handler, but not in the adjacent Slack appmention eventhandler.A Slack...

PraisonAI recipe.run_stream skips dangerous-tool policy enforcement

PraisonAI recipe.runstream skips dangerous-tool policy enforcement SummaryPraisonAI recipe execution blocks default-denied dangerous tools unless thecaller explicitly passes allowdangeroustools=True. The normal recipe.runpath enforces this with checktoolpolicy. The streaming path,recipe.runstream...

PraisonAI ToolsMCPServer legacy SSE transport accepts attacker Host/Origin and exposes registered tools

PraisonAI ToolsMCPServer legacy SSE transport accepts attacker Host/Origin and exposes registered tools Summarypraisonaiagents.mcp.ToolsMCPServer.runsse builds a Starlette MCPHTTP+SSE server around mcp.server.sse.SseServerTransport. The server exposes/sse and /messages/, but it does not validate...

bytedance InfiniStore: Denial of Service via Non-Cryptographic Hashing in InfiniStore KV Map

A vulnerability was found in bytedance InfiniStore up to 0.2.33. The impacted element is the function purgekvmap in the library /src/infinistore.h of the component KV Map Handler. Performing a manipulation results in inefficient algorithmic complexity. The attack requires a local approach. The...

ansible-core: Argument injection in ansible-galaxy role install leads to arbitrary code execution

A flaw was found in ansible-core. The ansible-galaxy role install command processes dependency specifications from a role's meta/requirements.yml file. Due to improper neutralization of argument delimiters, a malicious role author can inject arbitrary git configuration flags through the src field...

LMCache: 16-bit multimodal hash collision can poison KV cache entries

A flaw has been found in LMCache up to 0.4.6. This affects the function hexhashtoint16 of the file lmcache/integration/vllm/utils.py of the component KV Cache Handler. Executing a manipulation can lead to use of weak hash. The attack needs to be launched locally. The attack requires a high level ...

GPTCache: File and image cache keys collide because BufferedReader.peek() only reads the buffered prefix

A vulnerability was detected in zilliztech GPTCache up to 0.1.44. Affected by this issue is the function BufferedReader.peek of the file gptcache/processor/pre.py of the component Cache Key Handler. Performing a manipulation of the argument inputdata"image" results in use of weak hash. The attack...

mlrun: DataFrame hash collisions can cause dataset artifact path conflicts and silent data corruption

A vulnerability has been found in mlrun up to 1.12.0-rc3. This impacts the function mlrun.utils.helpers.calculatedataframehash of the file mlrun/utils/helpers.py of the component DataFrame Hash Handler. The manipulation leads to use of weak hash. The attack can only be performed from a local...

PYSEC-2026-3452

Pillow is a Python imaging library. From 12.0.0 through 12.2.0, Pillow's EPS parser in PIL/EpsImagePlugin.py accepts a negative byte count in the %%BeginBinary directive, allowing a crafted EPS file to cause Image.open to seek backwards to the same directive and parse it repeatedly in an infinite...

`lxml_html_clean.Cleaner` does not strip `javascript:` URLs from namespaced URL attributes

lxmlhtmlclean.Cleaner does not strip javascript: URLs from namespaced URL attributes xlink:hrefReporter: Guillem Lefait · Date: 2026-05-10Affected: lxml ≤ 6.1.0 and lxmlhtmlclean ≤ 0.4.4 latest stableConfirmed against: lxml 6.1.0 + lxmlhtmlclean 0.4.4 on Python 3.13.5, 3.14.4, and 3.15.0a8 libxml...

Open WebUI allows limited stored XSS vila uploaded html file

SummaryLow privileged users can upload HTML files which contain JavaScript code via the /api/v1/files/ backend endpoint. This endpoint returns a file id, which can be used to open the file in the browser and trigger the JavaScript code in the user's browser. Under the default settings, files...

Langroid: SQLChatAgent dangerous-function blocklist can be bypassed with quoted or schema-qualified pg_read_file calls

SQLChatAgent validatequery dangerous-pattern regex is bypassable via quoted/commented/qualified function names SummaryThe SQLChatAgent SQL-injection mitigation, with default allowdangerousoperations=False, combines a raw-text regex blocklist DANGEROUSSQLPATTERNS with a sqlglot SELECT-only stateme...

Langroid: SQLChatAgent _validate_query blocklist misses pg_read_file family enabling arbitrary file read

SummarySQLChatAgent in langroid ships a validatequery defense-in-depth layerwhose DANGEROUSSQLPATTERNS regex blocklist enumerates dangerous SQLprimitives by specific function name. The list misses the canonicalPostgreSQL filesystem-disclosure family pgreadfile, pgstatfile,pglslogdir, pglswaldir,...

Open Babel has out-of-bounds write in MOPAC translationVectors[] (UNIT CELL TRANSLATION)

SummaryA memory-safety vulnerability in Open Babel's MOPAC output parserallowed an out-of-bounds write into the translationVectors arraywhen reading the "UNIT CELL TRANSLATION" block of a crafted inputfile. DetailsThe MOPAC output reader stored translation vectors from the UNIT CELLTRANSLATION...

PGHoard: Password written to debug log

ImpactWhen using .pgpass, database connection information including the username and password will be logged at the debug level. PatchesUpgrade to version 2.7.1 or greater. WorkaroundsFilter out debug-level logs. ReferencesThis issue was discovered by BugCrowd user DRAKOKORIAN...

Open WebUI BOLA: `search_knowledge_files` Allows Unauthorized Knowledge Base File Enumeration

SummaryOpen WebUI has a Broken Object Level Authorization BOLA vulnerability in the builtin searchknowledgefiles tool.When native function calling is enabled and the selected model has no attached knowledge bases, an authenticated user can call searchknowledgefiles with an arbitrary knowledgeid...

Tornado has out-of-bounds memory access via C extension

SummaryTornado's optional native extension tornado.speedups implements websocketmask without validating that the mask argument is exactly four bytes long. The C function reads four bytes from mask unconditionally, even when Python passes a shorter byte string. This can read beyond the provided...

praisonai-platform: Any workspace member can rewrite workspace name, description, and settings via PATCH /workspaces/{id}

SummaryType: Authorization bypass enabling workspace metadata + settings tampering. The PATCH /workspaces/workspaceid endpoint is gated only by requireworkspacememberworkspaceid default minrole="member". Any member can rewrite the workspace's name, description, and the settings JSON blob. The...

praisonai-platform: list_issue_activity returns activity log for any issue regardless of workspace ownership

SummaryType: Insecure Direct Object Reference. The GET /workspaces/workspaceid/issues/issueid/activity endpoint is gated by requireworkspacememberworkspaceid and dispatches to ActivityService.listforissueissueid, which executes SELECT FROM activity WHERE issueid = :issueid with no workspace...

PraisonAI vulnerable to unauthenticated arbitrary file read via MCP workflow.show, workflow.validate, deploy.validate

SummaryThe fix for GHSA-9mqq-jqxf-grvw / CVE-2026-44336 is incomplete. The original advisory description named four vulnerable handlers in mcpserver/adapters/clitools.py: "registers four file-handling tools by default, praisonai.rules.create, praisonai.rules.show, praisonai.rules.delete, and...

compliance-trestle Vulnerable to SSRF in Remote Fetching Subsystem

A source code audit led to the discovery of three significant security vulnerabilities in the trestle/core/remote/cache.py module.Finding 1 Critical: SSRF CWE-918The HTTPSFetcher.dofetch method passes a user-supplied URL directly to requests.get without validation. This allows an attacker to...

AsyncSSH `AuthorizedKeysFile %u` path traversal allows attacker-selected authorized keys to authenticate a traversal username

SummaryAsyncSSH 2.22.0 expands the OpenSSH-compatible AuthorizedKeysFile %u token with the raw SSH username during pre-authentication server config reload. A server configured with a documented per-user key pattern such as AuthorizedKeysFile authorizedkeys/%u can be made to read an authorized-key...

Flask-Security-Too OAuth reauthentication freshness bypass via cross- user OAuth identity acceptance

Summary Flask-Security-Too 5.8.0's OAuth reauthentication flow can mark a session as fresh after verifying an OAuth account that belongs to a different user. If an attacker can operate an already-authenticated but stale victim session, they can complete OAuth verification using their own OAuth...

Crawlee for Python: SSRF via sitemap-derived URLs

Overview- Vulnerability type: Blind SSRF- Affected components: src/crawlee/utils/sitemap.py, src/crawlee/utils/robots.py, src/crawlee/requestloaders/sitemaprequestloader.py, and all built-in HTTP clients.- Trigger: an attacker-controlled sitemap or robots.txt containing a URL that points to an...

pyload-ng: SSRF via HTTP Redirect Bypass in parse_urls API

SummaryThe SSRF mitigation added in commit 33c55da for GHSA-7gvf-3w72-p2pg is incomplete. The PREREQFUNCTION-based private IP check was correctly applied to HTTPChunk download path but not to HTTPRequest used by the parseurls API. An authenticated attacker can supply a URL pointing to an...

Pipecat: Path Traversal in Pipecat Runner `/files` Endpoint — Arbitrary File Read via `%2F`-Encoded Separator

SummaryA path traversal vulnerability exists in Pipecat's development runner src/pipecat/runner/run.py. When the runner is started with the --folder flag, it exposes a GET /files/filename:path download endpoint. The filename path parameter is concatenated directly onto args.folder with no...