Juillet 2026

thumbor

Thumbor is an open-source photo thumbnail service by globo.com. Prior to 7.8.0, Thumbor’s HMAC validation can be bypassed due to the use of Python’s .replace() when removing the signature from the URL before validation. Since .replace() removes all occurrences of the substring, an attacker can insert the same signature multiple times in the URL and manipulate the final URL used for validation. This allows crafting URLs where the validated string differs from the actual requested resource, enabling loading images from unintended domains or paths. This issue is fixed in 7.8.0.

masci banks

Banks generates meaningful LLM prompts using a simple template language. In versions prior to 2.4.3, banks parses Tool JSON objects from the rendered body of {% completion %} blocks and later resolves their import_path field through importlib.import_module(...) + getattr(...) to obtain the callable that handles a tool call. There is no allowlist or sanitization on import_path, so any importable Python attribute (e.g. os.system, subprocess.getoutput) can be selected. When the LLM emits a tool_calls entry whose function.name matches the attacker-supplied tool name, the resolved callable is invoked with kwargs decoded from tool_call.function.arguments, yielding arbitrary code execution in the banks-hosting process. This is distinct from GHSA-gphh-9q3h-jgpp / CVE-2026-44209. That advisory was fixed in 2.4.2 by switching src/banks/env.py from Environment to SandboxedEnvironment. The fix does not touch src/banks/extensions/completion.py, and the unsafe import + getattr chain still executes on 2.4.2. The malicious Tool JSON is plain text in the rendered template body — it requires no Jinja attribute access, so the sandbox is irrelevant. This issue has been fixed in version 2.4.3.

masci banks

Banks generates meaningful LLM prompts using a simple template language. In versions prior to 2.4.4, all four media filters (image, audio, video, document) in banks accept untrusted user input as file paths via Path(value) and pass them directly to open(file_path, "rb") without any path sanitization, canonicalization, or directory restriction. An attacker who controls template variables passed to a banks Prompt can use path traversal (../) to read arbitrary files accessible to the Python process—including .env files, SSH keys, cloud credentials, source code, /etc/passwd, and /etc/shadow—with the content returned base64-encoded in the rendered prompt output, making exfiltration trivial. This is particularly dangerous for applications that use banks to process user-provided template variables before sending prompts to an LLM. This issue has been fixed in version 2.4.4.

Eclipse Foundation Eclipse CSI - PIA

PIA's `POST /v1/upload/sbom` endpoint accepts a Bearer JWT and checks its **unverified** `iss` claim against an issuer allowlist using Python's `urlparse` before performing OIDC discovery with `requests`. Because `urlparse` and `requests`/`urllib3` parse an authority string containing a backslash (e.g. `https://attacker-host\@ci.eclipse.org/`) into *different* hostnames, an attacker can craft an issuer that passes the allowlist check yet drives `requests` — and subsequently `urllib.request.urlopen` for JWKS retrieval — to connect to an arbitrary attacker-chosen host, port, and scheme.

Linuxfabrik monitoring-plugins

Linuxfabrik monitoring-plugins provides Python monitoring plugins for Icinga, Nagios, and related monitoring systems. In 6.0.0 and earlier, the redfish-* plugins built request URLs by concatenating an operator-supplied base URL with response-supplied @odata.id links, allowing a malicious or compromised BMC to redirect authenticated Redfish requests and disclose X-Auth-Token or HTTP Basic credentials.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Pydantic v2 models, dataclasses, TypedDict, and msgspec.Struct from OpenAPI, JSON Schema, GraphQL, Avro, Protobuf, and raw JSON, YAML, or CSV. From 0.11.6 until 0.64.0, datamodel-code-generator allows attacker-controlled x-python-import or customTypePath schema extensions to reach src/datamodel_code_generator/parser/jsonschema.py and generated import handling through Import.from_full_path and Imports.create_line in src/datamodel_code_generator/imports.py, allowing a newline to break out of an import statement and execute Python code when the generated model is imported. This issue is fixed in version 0.64.0.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Python data models from schema definitions. From 0.59.0 until 0.62.0, XML Schema parsing in src/datamodel_code_generator/parser/xmlschema.py for --input-file-type xmlschema resolves xs:include, xs:import, xs:redefine, and xs:override schemaLocation values outside the input base path, allowing arbitrary local files to be read and reflected into generated models. This issue is fixed in version 0.62.0.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Python data models from schema definitions. From 0.9.1 until 0.61.0, src/datamodel_code_generator/http.py http.get_body accepts --url targets and redirect chain targets without host/IP validation, allowing server-side request forgery against loopback, private, link-local, metadata, and other network-accessible resources. This issue is fixed in version 0.61.0.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Pydantic v2 models, dataclasses, TypedDict, and msgspec.Struct from OpenAPI, JSON Schema, GraphQL, Avro, Protobuf, and raw JSON, YAML, or CSV. From 0.52.1 until 0.60.2, datamodel-code-generator interpolates validators from --extra-template-data in src/datamodel_code_generator/model/pydantic_v2/base_model.py through _process_validators into @field_validator decorators without safe validation, allowing Python code execution when the generated Pydantic v2 model is imported. This issue is fixed in version 0.60.2.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Python data models from schema definitions. From 0.51.0 until 0.60.2, x-python-type values parsed by src/datamodel_code_generator/parser/jsonschema.py in _get_python_type_override are inserted into generated field annotations without sufficient validation, allowing attacker-controlled JSON Schema content to execute Python code when the generated module is imported. This issue is fixed in version 0.60.2.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Python data models from schema definitions. From 0.14.1 until 0.60.2, the --extra-template-data comment field is rendered into Python comments in src/datamodel_code_generator/model/template/TypeAliasAnnotation.jinja2, src/datamodel_code_generator/model/template/TypedDict.jinja2, src/datamodel_code_generator/model/template/dataclass.jinja2, src/datamodel_code_generator/model/template/msgspec.Struct.jinja2, src/datamodel_code_generator/model/template/pydantic/BaseModel.jinja2, and src/datamodel_code_generator/model/template/pydantic_v2/BaseModel.jinja2 without neutralizing carriage returns in Python # comments, allowing an attacker-controlled comment value to inject Python code into generated models that runs when imported. This issue is fixed in version 0.60.2.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Pydantic v2 models, dataclasses, TypedDict, and msgspec.Struct from OpenAPI, JSON Schema, GraphQL, Avro, Protobuf, and raw JSON, YAML, or CSV. From 0.17.0 until 0.60.2, datamodel-code-generator preserves attacker-controlled default_factory values in src/datamodel_code_generator/parser/jsonschema.py through JsonSchemaObject.init and get_field_extras and emits them into Field(default_factory=...) or field(default_factory=...), allowing Python expression execution when the generated model is imported. This issue is fixed in version 0.60.2.

koxudaxi datamodel-code-generator

datamodel-code-generator generates Python data models from schema definitions. Prior to 0.60.1, GraphQL Union description values in src/datamodel_code_generator/model/template/UnionTypeStatement.jinja2 and src/datamodel_code_generator/model/template/UnionTypeStatement.py312.jinja2 are rendered into Python comments without neutralizing carriage returns in Python # comments, allowing attacker-controlled GraphQL schema content to inject Python code into generated models that runs when imported. This issue is fixed in version 0.60.1.

nessshon tonapi

pytonapi is a Python SDK for TONAPI that provides REST API, streaming, and webhook access to the TON blockchain. From 2.0.0 to 2.2.0, TonapiWebhookDispatcher fails to validate the Authorization header when a webhook handler is registered with the documented path argument, because setup() stores bearer tokens only under the default suffix paths and never adds the custom path to the token map, so self._tokens.get(path) returns None and the authentication guard is skipped. An unauthenticated remote attacker can POST forged payloads to the custom webhook endpoint and trigger victim-defined handlers. This issue is fixed in version 2.2.1.

tomaka rouille

Rouille 0.3.3 through 3.6.2 contains an HTTP request smuggling vulnerability that allows remote attackers to bypass access controls by injecting bare line feed characters (0x0A) into client-supplied request header values that are copied verbatim to upstream connections without validation. Attackers can craft a header value containing a complete additional HTTP request that is interpreted as a separate request by backends such as Go net/http and Python http.server, causing the backend to process a smuggled request with attacker-chosen method, path, and headers that bypasses the rouille handler's access control logic.

Apache Software Foundation Apache Thrift · Apache thrift

Improper Handling of Highly Compressed Data (Data Amplification) vulnerability in Apache Thrift C++, Java, Python, Go, D, C/GLib bindings. This issue affects Apache Thrift: before 0.24.0. Users are recommended to upgrade to version 0.24.0, which fixes the issue.

Apache Software Foundation Apache Thrift · Apache thrift

Loop with Unreachable Exit Condition ('Infinite Loop') vulnerability in Apache Thrift Python, Go, PHP and Java bindings.This issue affects Apache Thrift: before 0.24.0. Users are recommended to upgrade to version 0.24.0, which fixes the issue.

Apache Software Foundation Apache Thrift · Apache thrift

Improper Handling of Highly Compressed Data (Data Amplification) vulnerability in Apache Thrift Python bindings. This issue affects Apache Thrift: before 0.24.0. Users are recommended to upgrade to version 0.24.0, which fixes the issue.

codexu NoteGen · Notegen notegen

NoteGen before 0.32.0 grants the Tauri shell plugin shell:allow-execute capability for bash, python, and python3 with arbitrary arguments in the default desktop capabilities. JavaScript running in the application webview can therefore invoke plugin:shell|execute to run attacker-controlled operating system commands with the privileges of the NoteGen process. In combination with script execution in the webview (for example via chat XSS), this enables full remote code execution on the user's machine.

koxudaxi datamodel-code-generator

datamodel-code-generator prior to version 0.70.0 contains a code injection vulnerability that allows attackers who control input schemas to achieve remote code execution by supplying a malicious customBasePath value containing embedded newlines and a dot-free Python expression. The crafted value is emitted verbatim into a generated 'from ... import ...' statement without identifier validation, causing arbitrary Python code to execute when the generated module is imported.

ntlk · Nltk nltk

NLTK (Natural Language Toolkit) before version 3.9.3 contains an eval injection vulnerability in the nltk.collocations module that allows an attacker who controls command-line arguments to execute arbitrary Python code. When collocations.py is invoked directly, the __main__ block passes command-line arguments directly to eval() as suffixes of BigramAssocMeasures without allowlist validation or sanitization, enabling an attacker to supply a Python expression that escapes the intended attribute lookup and executes arbitrary code including OS commands via the os module.

OpenStack Ironic Python Agent

In OpenStack Ironic Python Agent through 11.6.0, a project-scoped user with the manager role can achieve arbitrary code execution on a running Ironic-Python-Agent via a maliciously constructed configuration, because the value of ntp_server is passed to a shell.

AWS bedrock-agentcore 1.18.1

Improper neutralization of argument delimiters in the install_packages() method in AWS Bedrock AgentCore Python SDK before 1.18.1 might allow a remote authenticated user to execute arbitrary commands within the Code Interpreter sandbox via crafted package name arguments. To mitigate this issue, users should upgrade to the patched version 1.18.1.

JetBrains PyCharm

In JetBrains PyCharm before 2026.1.4, 2026.2 arbitrary code execution via malicious Python executable was possible on untrusted project open

Red Hat Ansible Automation Platform 2

A flaw was found in the Ansible Lightspeed Visual Studio Code extension. This Command Injection vulnerability (CWE-78) allows a remote attacker to execute unauthorized commands on a user's system. The issue occurs because the `ansible.python.activationScript` setting, intended for a virtual environment activation script, does not properly validate user input as a file path. If a user opens or executes a specially crafted project, an attacker could exploit this to gain complete control over the user's system with the privileges of the Visual Studio Code application.

home-assistant Home Assistant Core

Home Assistant Core before 2026.7.0 contains a path traversal vulnerability in the backup-restore function that allows attackers to write files to arbitrary absolute filesystem paths by supplying a crafted tar archive with a SYMTYPE entry containing a benign member name paired with an absolute linkname pointing outside the extraction directory. Because the official Docker image runs the Home Assistant process as root and the subsequent regular-file entry is written through the unvalidated symlink, attackers can achieve remote code execution by overwriting auto-imported Python paths such as site-packages/sitecustomize.py or custom component directories.

Extreme Networks Switch Engine (EXOS)

ExtremeXOS (EXOS) uses a challenge-response mechanism to authorize access to the privileged debug-mode function. The challenge value is generated using an insufficiently random source, which under certain conditions may allow an attacker to predict the expected response and activate debug-mode without authorization. Depending on device configuration and version, this may enable escalation to root-level access and persistent modification of the device software stack. Exploitation requires either a valid low-privilege account on the device (remote scenario) or physical serial console access (local scenario). This vulnerability is distinct from CVE-2017-14329, which addressed a different issue involving Python script privileges. Extreme would like to thank Hadrien Barral (Université Gustave Eiffel) and Georges-Axel Jaloyan (French Ministry of the Interior) for responsible disclosure of their findings.

meshtastic firmware · Meshtastic meshtastic firmware

Meshtastic is an open source mesh networking solution. Prior to version 2.7.23.b246bcd, a single node advertising a User.long_name that contains a malformed character encoding can render other radios unusable over BLE when managed through the iOS app. The malformed name does not need to be maliciously crafted — it can arise from ordinary buffer truncation and has been observed occurring naturally in the wild. At least one code path could place a null terminator in the middle of a multibyte sequence, leaving a malformed User.long_name in the node database. The problem surfaced downstream: the iOS app enforced encoding validation and therefore cannot parse a node database once it contains a poisoned entry. This caused BLE sync to enter a fail/retry loop, resulting in loss of control over the affected device. For a typical user managing their radio with the iOS app, the device becomes effectively unusable until the poisoned node ages out of the on-device database, or unless they have an alternate management path (e.g., the Python CLI, which can be used to identify and remove the offending entries manually). Because the malformed name propagates through the mesh, the temporary presence of a single affected node can degrade BLE management for iOS users across a wide geographical area for an extended period. Less technical users have no straightforward recovery path. Starting in version 2.7.23.b246bcd, the firmware has added input sanitization and regression tests demonstrating recovery for already-poisoned devices. The apps have also taken steps to ensure more graceful handling of malformed encoding sequences as well.

urwid

The urwid web display backend (urwid/display/web.py) generates web session identifiers (urwid_id) in Screen.start() by concatenating two random.randrange(10**9) calls that use Python's Mersenne Twister PRNG, which is not cryptographically secure. Each call consumes approximately 30 bits of PRNG state, and the Mersenne Twister internal state is approximately 19,937 bits, so an attacker who observes approximately 334 session IDs (for example via the X-Urwid-ID HTTP response header) can fully reconstruct the internal state and predict all past and future session IDs (Path B). The same identifier is also used as the filename of a FIFO created in the world-listable /tmp directory (for example /tmp/urwid375487765176907690.in), so any local user on the host can list /tmp to enumerate active session tokens directly (Path A). With a valid session ID, an attacker can read the victim's terminal screen via the polling endpoint, inject keystrokes into the victim's session (yielding OS-level code execution with the session owner's privileges if the session runs a shell), and inject exit sequences or flood the FIFO to terminate or crash the session. A prior Bandit S311 warning on this usage was suppressed with # noqa: S311 rather than fixed

scikit-hep uproot

uproot dynamically generates Python class source code from ROOT TStreamerInfo records in a file and compiles it at runtime. Some file-controlled streamer metadata fields (for example, streamer element names) are interpolated into the generated Python source without safe quoting via repr() or the !r format specifier. An attacker who can supply a crafted ROOT file can place Python expression-breaking content into a streamer metadata field. When uproot generates and invokes the corresponding reader method, the injected Python expression is evaluated in the context of the process opening the file, resulting in arbitrary Python code execution in applications that open or process attacker-controlled ROOT files with affected uproot code paths.

DataDog dd-trace-py

Datadog dd-trace-py is the Datadog Python APM client. Prior to 4.8.2, Datadog tracing libraries that implement W3C baggage propagation parse incoming baggage HTTP headers without enforcing DD_TRACE_BAGGAGE_MAX_ITEMS or DD_TRACE_BAGGAGE_MAX_BYTES limits on the extract path. A remote, unauthenticated attacker can send a request whose baggage header contains an arbitrarily large number of comma-separated key-value pairs or a single very large value, causing unbounded CPU and memory consumption and enabling a remote denial of service against HTTP services with baggage propagation enabled. This issue is fixed in version 4.8.2.

authlib joserfc

joserfc is a Python library that provides an implementation of several JSON Object Signing and Encryption (JOSE) standards. Prior to 1.6.8, joserfc.jwt.decode accepts attacker-forged HMAC-signed tokens when the caller-supplied verification key is the empty string or None, because HMACAlgorithm.sign and HMACAlgorithm.verify in src/joserfc/_rfc7518/jws_algs.py pass the output of OctKey.get_op_key(...) to hmac.new(...) and OctKey.import_key in src/joserfc/_rfc7518/oct_key.py only emits a SecurityWarning for keys shorter than 14 bytes without rejecting zero-length input. This issue is fixed in version 1.6.8.

IBM Langflow OSS · Langflow langflow

IBM Langflow OSS 1.0.0 through 1.10.1 Langflow could allow an authenticated user to execute arbitrary commands with elevated privileges on the system due to improper validation of user supplied input in the Python Interpreter component.

ronf asyncssh · Asyncssh project asyncssh

AsyncSSH is a Python package which provides an asynchronous client and server implementation of the SSHv2 protocol on top of the Python asyncio framework. Prior to 2.23.0, AsyncSSH expands the OpenSSH-compatible AuthorizedKeysFile %u token in asyncssh/config.py, asyncssh/connection.py, asyncssh/auth_keys.py, and asyncssh/misc.py with the raw SSH username during pre-authentication server config reload, allowing a server configured with AuthorizedKeysFile authorized_keys/%u to read an authorized-keys file outside the intended directory when the SSH username contains /, \, or .. path traversal segments and authenticate with an attacker-selected key file. This issue is fixed in version 2.23.0.

microsoft kiota

Kiota is an OpenAPI based HTTP Client code generator. Prior to 1.29.1 and 1.32.0, Kiota's Python generator let attacker-controlled enum value descriptions from x-ms-enum.values[].description flow through KiotaBuilder.SetEnumOptions into Documentation.DescriptionTemplate and PythonConventionService.RemoveInvalidDescriptionCharacters without newline sanitization, allowing generated inline comments to split and execute attacker-controlled Python code at module scope when generated modules were imported. This issue is fixed in version 1.29.1 and 1.32.0.

Lookyloo PlaywrightCapture

PlaywrightCapture stored capture-specific configuration and runtime data as mutable class-level variables rather than instance-level variables. Consequently, multiple Capture objects running within the same Python process could share state, including HTTP headers, cookies, browser storage, HTTP credentials, proxy configuration, user-agent settings, geolocation information, and captured request data. In a multi-user or concurrent deployment, information supplied during one capture could therefore persist and be reused by a subsequent or parallel capture. This could result in the disclosure of authentication cookies, credentials, browser storage, or captured request data belonging to another user. It could also cause requests to be performed with another capture's authentication context, headers, or proxy configuration, potentially enabling unauthorized access to remote resources or interference with other capture operations. The vulnerability is resolved by initializing all capture-specific settings and request data as instance variables in the Capture constructor, ensuring that state is isolated between capture operations.

modelcontextprotocol python-sdk · Lfprojects mcp python sdk

The MCP Python SDK, called mcp on PyPI, is a Python implementation of the Model Context Protocol (MCP). Prior to 1.28.1, the deprecated mcp.server.websocket.websocket_server transport accepted WebSocket handshakes without applying Host or Origin header validation, leaving no SDK-level way to restrict which origins could connect to applications that exposed that transport. This issue is fixed in version 1.28.1.

modelcontextprotocol python-sdk · Lfprojects mcp python sdk

The MCP Python SDK, called mcp on PyPI, is a Python implementation of the Model Context Protocol (MCP). From 1.23.0 until 1.27.2, default handlers installed by server.experimental.enable_tasks() for tasks/list, tasks/get, tasks/result, and tasks/cancel operate only on task identifiers without recording the session that created each task, allowing any connected client to enumerate, read results from, consume messages for, or cancel other clients' tasks. This issue is fixed in version 1.27.2.

modelcontextprotocol python-sdk · Lfprojects mcp python sdk

The MCP Python SDK, called mcp on PyPI, is a Python implementation of the Model Context Protocol (MCP). Prior to 1.27.2, the SSE and stateful Streamable HTTP transports mcp.server.sse.SseServerTransport and mcp.server.streamable_http_manager.StreamableHTTPSessionManager route requests to existing sessions using only the session_id query parameter or Mcp-Session-Id header without verifying the authenticated principal that created the session, allowing a different bearer-token-authenticated client with a known session ID to inject JSON-RPC messages into that session. This issue is fixed in version 1.27.2.

MervinPraison PraisonAI

PraisonAI (praisonaiagents) before 1.6.78 contains a remote code execution vulnerability in the plugin manager, which loads and executes arbitrary Python (.py) files from project-level and user-home .praisonai/plugins/ directories using importlib spec_from_file_location() and exec_module() without code signing, integrity verification, or sandboxing. An attacker who can write a malicious .py file to a plugin directory (for example via path traversal, a supply chain attack, or a compromised dependency) achieves arbitrary code execution when the plugin system initializes.

MervinPraison PraisonAI

PraisonAI before 4.6.78 contains a remote code execution vulnerability in JobWorkflowExecutor._exec_inline_python() due to insufficient AST validation of workflow script steps. Attackers can create malicious YAML workflow files with import os statements followed by os.system() calls that bypass sandbox checks and execute arbitrary OS commands with process privileges.

MervinPraison PraisonAI

PraisonAI before 4.6.78 fails to safely encode deployment configuration values when generating Python source code for API servers. Attackers can inject arbitrary Python expressions through the deploy.api.host and agents_file configuration parameters that execute when the generated server starts or handles requests.

tornadoweb tornado

Tornado is a Python web framework and asynchronous networking library. Prior to 6.5.6, Tornado gzip decompression routines processed limited-size chunks but did not enforce an overall limit on accumulated decompressed chunks, allowing a malicious server accessed by SimpleAsyncHTTPClient or an HTTPServer configured with decompress_request=True to consume effectively unlimited memory. This issue is fixed in version 6.5.6.

tornadoweb tornado

Tornado is a Python web framework and asynchronous networking library. Prior to 6.5.6, SimpleAsyncHTTPClient shallow-copied redirected requests and removed only the Host header, leaving Authorization, auth_username, auth_password, and auth_mode in place when a redirect changed scheme, host, or port. This issue is fixed in version 6.5.6.

pyasn1 · Pyasn1 pyasn1

pyasn1 is a generic ASN.1 library for Python. Prior to 0.6.4, the univ.Real type converted its mantissa, base, and exponent value to a Python float using exact big-integer exponentiation. A BER, CER, or DER encoded REAL value only a few bytes long can carry a very large exponent, causing float conversion through prettyPrint(), str(), comparison, arithmetic, int(), or an explicit float() call to consume excessive CPU and memory and hang applications that decode untrusted ASN.1 data and then print, log, or compare decoded objects. This issue is fixed in version 0.6.4.

pyasn1 · Pyasn1 pyasn1

pyasn1 is a generic ASN.1 library for Python. Prior to 0.6.4, the BER, CER, and DER decoders process OBJECT IDENTIFIER and RELATIVE-OID values in quadratic time relative to the number of arcs, so a small crafted payload containing an OID with many arcs consumes excessive CPU per decode() call and can deny service to applications that decode untrusted ASN.1 data. The corresponding encoders have the same quadratic behavior when an application re-encodes previously decoded attacker-supplied values. This issue is fixed in version 0.6.4.

pyasn1 · Pyasn1 pyasn1

pyasn1 is a generic ASN.1 library for Python. Prior to 0.6.4, the BER decoder shared by the CER and DER codecs parses long-form tags by accumulating continuation octets without an upper bound on the tag ID size, allowing a crafted input to force construction of an arbitrarily large integer with CPU cost growing quadratically and to trigger unhandled ValueError exceptions in Python 3.11+ error formatting paths. Any application decoding untrusted BER, CER, or DER input is affected. This issue is fixed in version 0.6.4.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. From 5.1.0 until 12.3.0, PdfParser.PdfStream.decode() in PIL/PdfParser.py calls zlib.decompress() with bufsize set to the PDF stream Length field without bounding the decompressed output size, allowing a crafted FlateDecode PDF stream to exhaust memory from a small file. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, Pillow's public rank-filter API can trigger a native heap out-of-bounds write when given a very large odd filter size because ImageFilter.RankFilter.filter() calls image.expand(size // 2, size // 2) before rank-filter size validation and ImagingExpand() computes output dimensions with unchecked signed int arithmetic. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, Pillow's ImageCms.ImageCmsTransform.apply(im, imOut) API can trigger controlled native heap corruption when the caller supplies an output image whose mode does not match the transform's declared output mode. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. From 8.2.0 through 12.2.0, src/libImaging/Jpeg2KDecode.c accumulates total_component_width across every tile in a JPEG2000 image instead of recomputing it per tile, allowing a crafted tiled JPEG2000 file to force substantially higher transient memory usage and trigger out-of-memory failures during decoding. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

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 loop. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, Pillow public image coordinate APIs can trigger a native heap out-of-bounds write when given coordinates near the signed 32-bit integer limits in Image.paste(), Image.crop(), or Image.alpha_composite(). This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. From 5.2.0 until 12.3.0, Pillow's TGA RLE encoder reads past its packed row buffer when saving a mode 1 image with TGA RLE compression, allowing adjacent process heap bytes to be copied into the generated TGA file. This issue is fixed in version 12.3.0.

gristlabs grist-core

Grist is spreadsheet software using Python as its formula language. Prior to 1.7.15, Grist contained two cross-site scripting vulnerabilities where an attacker-controlled value reached a link's href without scheme validation, so a javascript URL could run in a victim's Grist origin on a single click. On the account-selection page, /welcome/select-account used its next query parameter as the account buttons' link target. In document tours, the GristDocTour table's Link_URL column became a clickable button, allowing an editor of a shared document to store a javascript URL there that ran when another user opened the document and clicked the tour link. Because the script runs in the victim's authenticated session, it can call Grist APIs as the victim, reading or modifying data and changing sharing settings and access rules. A document editor could therefore escalate to owner-level access. This issue is fixed in version 1.7.15.

gristlabs grist-core

Grist is spreadsheet software using Python as its formula language. Prior to 1.7.15, several server-rendered Grist pages embedded user-controlled values into the page and into inline scripts without fully escaping them, allowing cross-site scripting. On the main application page, a document's name or description, set by a document editor, is rendered into the page that other users load when opening the document. On the OAuth2 end-of-flow page, the openerOrigin request parameter was reflected back into the served page. Injected script runs in the victim's Grist origin and can act through the authenticated session, reading or modifying data and changing sharing settings and access rules. A document editor could therefore escalate to owner-level access. This issue is fixed in version 1.7.15.

MervinPraison PraisonAI

PraisonAI (pip package praisonaiagents) before 1.6.78 contains an unsafe dynamic module loading vulnerability in AgentFlow._resolve_pydantic_class (src/praisonai-agents/praisonaiagents/workflows/workflows.py). When a workflow step uses a string output_pydantic reference, the framework locates and imports a sibling tools.py from the workflow file's directory via importlib exec_module without sandboxing, ignoring the PRAISONAI_ALLOW_*_TOOLS environment variables. An attacker who controls a workflow file and its sibling tools.py can execute arbitrary Python code with the workflow runner's privileges when the workflow is executed via WorkflowManager or after load_yaml.

jg-rp liquid

Python Liquid is a Python engine for the Liquid template language. Prior to 2.2.1, given a malformed {% case %} tag without an associated {% when %} or {% else %} block and no terminating {% endcase %} tag, Python Liquid hangs in an infinite loop at parse time because liquid.TokenStream.eof did not give the EOF token matching kind and value fields, allowing malicious template authors to craft templates for a denial of service attack. This issue is fixed in version 2.2.1.

Python Software Foundation CPython · Python python

The incremental HTML parser (html.parser.HTMLParser) allows for CPU denial-of-service through repeated unterminated markup declarations when processing uncontrolled data.

stanfordnlp stanza · Stanford stanza

Stanza is a Stanford NLP Python library for tokenization, sentence segmentation, NER, and parsing of many human languages. Prior to 1.12.2, Stanza model loaders such as stanza.models.common.pretrain.Pretrain.load() attempt torch.load(..., weights_only=True) but fall back to torch.load(..., weights_only=False) on attacker-controllable pickle.UnpicklingError, allowing a malicious .pt pretrain or model file to execute arbitrary pickle code when a Stanza NLP pipeline loads it. This issue is fixed in version 1.12.2.

zopefoundation RestrictedPython

RestrictedPython is a tool that helps to define a subset of the Python language which allows to provide a program input into a trusted environment. Prior to 8.3, check_function_argument_names() rejected protected guard hook names for regular, variadic, and keyword-only arguments but omitted positional-only arguments, allowing __getattr__, _getitem_, _write_, or _print_ to be shadowed by a local parameter and bypass the embedding application's access policy. This issue is fixed in version 8.3.

miurahr py7zr

py7zr is a Python-based library and utility to support 7zip archive compression, decompression, encryption and decryption. Prior to 1.1.3, PackInfo._read() in archiveinfo.py used an O(n^2) cumulative sum pattern for attacker-controlled numstreams values parsed from archive headers, allowing a crafted .7z archive to cause excessive CPU consumption during SevenZipFile.init() before extraction. This issue is fixed in version 1.1.3.

miurahr py7zr

py7zr is a Python-based library and utility to support 7zip archive compression, decompression, encryption and decryption. Prior to 1.1.3, py7zr's Worker.decompress() extracted archive entries without tracking total decompressed size, allowing a crafted .7z file such as a 15.6 KB archive that expands to 100 MB to exhaust disk or memory before extraction completes. This issue is fixed in version 1.1.3.

ronf asyncssh

AsyncSSH is a Python package which provides an asynchronous client and server implementation of the SSHv2 protocol on top of the Python asyncio framework. Prior to 2.23.1, a malicious SSH server can write arbitrary files on the asyncssh SCP client's filesystem by sending filenames containing ../ traversal sequences because _parse_cd_args in scp.py returns server-provided names verbatim and _recv_files joins them to the destination path without enforcing the target directory boundary. This issue is fixed in version 2.23.1.

httplib2 · Httplib2 project httplib2

httplib2 is a comprehensive HTTP client library for Python. Prior to 0.32.0, httplib2 performs unbounded decompression of HTTP response bodies encoded with Content-Encoding: gzip or deflate in _decompressContent in httplib2/init.py, allowing a malicious or compromised HTTP server to return a small compressed payload that expands to an arbitrarily large size in memory and causes MemoryError or OOM-kill in the client process. This issue is fixed in version 0.32.0.

py-pdf pypdf · Pypdf project pypdf

pypdf is a free and open-source pure-python PDF library. Prior to 6.14.1, an attacker can craft a PDF with a page content stream containing a not terminated inline image, causing an infinite loop during inline image end marker detection such as when extracting page text. This issue is fixed in version 6.14.1.

py-pdf pypdf · Pypdf project pypdf

pypdf is a free and open-source pure-python PDF library. Prior to 6.14.2, an attacker can craft a PDF with a page content stream containing a not terminated inline image that uses the ASCII85 or ASCIIHex filters, causing an infinite loop during parsing such as when extracting page text. This issue is fixed in version 6.14.2.

BerriAI litellm · Litellm litellm

LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. Prior to 1.82.0-stable, LiteLLM's Custom Code Guardrails production create and update paths did not apply the same sandboxing and validation used by the test endpoint, allowing a privileged user with access to create or update guardrails to submit custom Python code that executed in the LiteLLM proxy environment and could expose secrets available to the process. This issue is fixed in version 1.82.0-stable.

py-pdf pypdf · Pypdf project pypdf

pypdf is a free and open-source pure-python PDF library. Prior to 6.14.0, an attacker can craft a PDF with repeated malformed cross-reference streams that cause pypdf to spend long runtimes recovering broken cross-reference table entries. This issue is fixed in version 6.14.0.

lepture mistune · Mistune project mistune

Mistune is a Python Markdown parser with renderers and plugins. Prior to 3.3.0, a Markdown document containing many repeated or distinct reference-link definitions causes quadratic work in src/mistune/block_parser.py and the ref_links environment dictionary handling, allowing denial of service through CPU exhaustion. This issue is fixed in version 3.3.0.

lepture mistune · Mistune project mistune

Mistune is a Python Markdown parser with renderers and plugins. Prior to 3.3.0, long sequences of well-formed double-asterisk or triple-asterisk emphasis pairs around a character cause quadratic work in src/mistune/inline_parser.py because the parser scans forward for matching close markers from every potential opening run, allowing denial of service in default Mistune parsing. This issue is fixed in version 3.3.0.

lepture mistune · Mistune project mistune

Mistune is a Python Markdown parser with renderers and plugins. Prior to 3.3.0, a run of closed tilde, equals-sign, or caret marker pairs around a character causes quadratic work in src/mistune/plugins/formatting.py when the strikethrough, mark, or insert plugin scans for matching markers from each possible start position, allowing denial of service through CPU exhaustion. This issue is fixed in version 3.3.0.

kovidgoyal calibre

calibre is an e-book manager. Prior to 9.10.0, a malicious EPUB, OPF, or PDF file can execute arbitrary Python code when its metadata is read by calibre, including through Add books or Edit books, by embedding a custom column definition with a python: template in calibre:user_metadata that is passed unsanitized to exec() in the template formatter. This issue is fixed in version 9.10.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, PIL/GdImageFile.py GdImageFile._open() read image dimensions from the GD 2.x header and stored them in self._size without calling Image._decompression_bomb_check(), allowing a crafted .gd file to trigger excessive C-heap allocation when loaded. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, PIL/BdfFontFile.py bdf_char() read the BBX width and height field from a BDF font file and passed attacker-controlled dimensions to Image.new() without calling Image._decompression_bomb_check(), bypassing Pillow's documented decompression bomb protection and allowing excessive memory allocation. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, PIL/FontFile.py FontFile.compile() assembled per-glyph images into a combined bitmap with Image.new("1", (xsize, ysize)) without calling Image._decompression_bomb_check(), allowing a font to trigger excessive allocation during conversion or saving. This issue is fixed in version 12.3.0.

python-pillow Pillow · Python pillow

Pillow is a Python imaging library. Prior to 12.3.0, PIL/PcfFontFile.py _load_bitmaps() read glyph dimensions from the PCF METRICS section and passed them directly to Image.frombytes() without calling Image._decompression_bomb_check(), allowing crafted PCF font data to cause excessive memory allocation. This issue is fixed in version 12.3.0.

trailofbits fickling

Trail of Bits fickling versions up to and including 0.1.10 do not include the Python standard library modules _posixsubprocess, site, and atexit in the UNSAFE_IMPORTS denylist (fickle.py). Because these modules are absent from the denylist, fickling's check_safety() function returns LIKELY_SAFE with zero findings for pickle payloads that invoke dangerous functions including _posixsubprocess.fork_exec (C-level process spawner capable of executing arbitrary binaries), site.execsitecustomize (executes arbitrary site customization code), and atexit._run_exitfuncs (triggers all registered exit handler callbacks). The fickling.load() API chains check_safety() into pickle.loads() as an explicit security gate; a LIKELY_SAFE verdict causes the payload to be deserialized and executed. This shares the same root cause as CVE-2026-22607 (cProfile), CVE-2025-67748 (pty), and CVE-2025-67747 (marshal/types). OvertlyBadEvals does not flag these modules because they are standard library imports. UnsafeImports does not flag them because they are not in the denylist. The UnusedVariables heuristic is defeated by the SETITEMS opcode pattern.

Picklescan

Picklescan before 0.0.33 fails to detect the numpy.f2py.crackfortran.getlincoef gadget in pickle __reduce__ methods, allowing arbitrary code execution. Attackers can craft malicious pickle files that execute arbitrary Python code when loaded, bypassing Picklescan's safety checks and enabling supply-chain poisoning of shared model files.

picklescan

picklescan before 0.0.29 fails to detect the built-in python profile.Profile.run function when used in pickle reduce methods, allowing attackers to execute arbitrary code. Remote attackers can craft malicious pickle files that bypass picklescan detection and achieve code execution upon deserialization.

picklescan

picklescan before 0.0.29 fails to detect the built-in Python trace.Trace.runctx function when used in pickle file reduce methods, allowing attackers to execute arbitrary code. Remote attackers can craft malicious pickle files with trace.Trace.runctx payloads that bypass picklescan detection and execute code upon pickle.load() invocation.

msgpack-python · Msgpack messagepack

MessagePack is the serializer implementation for Python msgpack.org. Prior to 1.2.1, there is an Out-of-bounds read/crash on Unpacker reuse after a caught error, potentially leading to a DoS attack. If the Unpacker is used repeatedly after an error occurs, the process may crash with a SEGV. This issue has been fixed in version 1.2.1.