download_wheel() runs pip download and hands the result straight to the seeder, with nothing checking the bytes it gets back. The embedded pip and setuptools wheels carry a BUNDLE_SHA256 that virtualenv checks on every load, but a wheel fetched over the network for the periodic-update feature or the –download flag had no equivalent verification. A compromised index, a stale mirror, or a MITM'd download (when TLS is intercepted, e.g. via …
pyvenv.cfg is a line-based format with no escape syntax. PyEnvCfg.write() wrote values verbatim, while PyEnvCfg._read_values() parses the file with str.splitlines(). A value containing a line boundary therefore became additional configuration lines, and because reading is last-wins, the injected keys replaced any key written earlier in the file.
urllib3 supports configuring TLS independently for an HTTPS proxy and the target server. proxy_ssl_context, proxy_assert_hostname, and proxy_assert_fingerprint configure the TLS connection to the proxy. ssl_context and the other target-specific TLS parameters configure the connection to the target server. In urllib3 versions 1.26.0 through 2.7.0, these configurations were not consistently separated. Depending on the proxy mode, urllib3 could: Ignore proxy_ssl_context and use the target server's SSL context for the TLS connection …
urllib3's streaming API is designed for efficiently handling large HTTP responses by reading the content in chunks, rather than loading the entire response body into memory at once. When decoding a chunked-transfer-encoded response, this API reads each chunk's size field by buffering until it sees \n or EOF. A malicious HTTP server can return Transfer-Encoding: chunked and then send a very long run of bytes without any newline, causing the …
urllib3's streaming API is designed for the efficient handling of large HTTP responses by reading content in chunks instead of loading the entire response body into memory at once. urllib3 can decompress response bodies according to the HTTP Content-Encoding header. When streaming a compressed, chunked response, urllib3 first consumes data already buffered by the decoder before reading the next HTTP chunk. However, urllib3 versions from 2.6.2 through 2.7.0 could enter …
HTTPServerRequest.init in tornado/httputil.py parses the URL query string via parse_qs_bytes() with no field-count limit — while the sibling POST-body parsing path (parse_body_arguments) received a max_num_fields=1000 cap added earlier in this exact same release (v6.5.8, commit 8d6363ed), explicitly to bound parsing cost for the identical underlying primitive. This leaves the query-string path with the resource-exhaustion exposure the body-path fix was meant to close. File: tornado/httputil.py, line 553 (HTTPServerRequest.init)
StaticFileHandler allows an unauthenticated attacker to read arbitrary files from the server's filesystem by requesting a path that resolves to a symbolic link placed inside the static root directory. Any application that serves user-uploadable content, or whose static directory is populated by a build/deploy pipeline that creates symlinks (e.g. npm link, webpack, Docker volume mounts, CDN sync tools), is affected. An attacker who can trigger the creation of a symlink …
An unbounded memory accumulation (decompression bomb) in tornado.curl_httpclient.CurlAsyncHTTPClient — the client-side sibling gap of CVE-2026-49855 — verified end-to-end on the 2026-08-15 master snapshot (6.6.dev1) and present unchanged in the latest release tag v6.5.8 and on master (checked 2026-08-17). When a Tornado application configures the curl client (the documented deployment for proxy support / advanced TLS options) and fetch()es an attacker-chosen or attacker-compromised URL with default decompress_response=True, a malicious server replying …
serialize-javascript escapes its output so it is safe to embed inside a <script> element. In 7.1.1 that guarantee does not hold for function values: a crafted function body can carry a literal, unescaped </script> into the output, terminating the script element early so the remainder is parsed as HTML. SCRIPT_CLOSE_REGEXP used </script[^>]*> as its first alternative. The character class excludes only >, so a single match could run from one …
A russh server can be driven to unbounded heap growth (process OOM / kill) by a peer that speaks only standard SSH messages, in the default configuration. The peer starts a key re-exchange (sends SSH_MSG_KEXINIT) but never sends the follow-up SSH_MSG_KEX_ECDH_INIT, leaving the server's kex state machine in SessionKexState::InProgress indefinitely. While a rekey is in progress the server's three message-drain paths are all gated off (if !self.kex.active()), but the network-read …
SshBlockCipher implementations (AES-CBC, AES-CTR, 3DES-CBC, etc.) report needs_mac() == true, meaning they are documented/intended to always be paired with a separate integrity MAC. However, key-exchange negotiation only checks needs_mac() inside the fallback branch of MAC algorithm selection (used when no common MAC algorithm exists). If both peers' preferred MAC lists simply contain none and it is successfully negotiated through the normal selection path, nothing rejects pairing none with a cipher …
The purpose of hybrid key exchange is defense-in-depth: if either the classical algorithm (X25519) or the post-quantum algorithm (ML-KEM 768) is broken, the combined secret remains secure. By injecting a zero X25519 public key, an attacker eliminates the classical contribution entirely, reducing security to depend solely on ML-KEM. This matters in two scenarios: If ML-KEM 768 is later found to have a weakness (the explicit threat model hybrid KEX is …
Russh: Configured server auth-attempt cap is not enforced in the USERAUTH_REQUEST runtime path
CVE-2026-68930 was fixed by adding Session::is_established_channel() in russh/src/server/encrypted.rs, which gates every channel-scoped SERVER-side message (CHANNEL_REQUEST, CHANNEL_DATA, CHANNEL_EOF, CHANNEL_CLOSE, CHANNEL_WINDOW_ADJUST, CHANNEL_EXTENDED_DATA) on enc.channels.get(&channel).is_some_and(|c| c.confirmed) before invoking any Handler callback. The identical validation was never added to the CLIENT side (russh/src/client/encrypted.rs), which processes channel-scoped messages sent by the SSH SERVER once the client has authenticated.
PyJWT.decode()/decode_complete() mutates a caller-supplied options dict in place whenever verify_signature is falsy, adding verify_exp/verify_nbf/verify_iat/verify_aud/verify_iss/verify_sub/verify_jti keys directly onto that object. If application code reuses the same options dict across calls (a config object, a module-level constant, a wrapper's self.options), a later call that explicitly sets verify_signature=True for a full, normal verification silently inherits the leftover verify_exp=False etc. from an earlier verify_signature=False call on the same object – and accepts an expired, …
PyJWKClient.get_signing_key_from_jwt(token) — the first step of the JWKS verification flow documented in docs/usage.rst — must decode a token's payload before its signature can be checked, via jwt.api_jwt.decode_complete(token, options={"verify_signature": False}). That call parses the payload with json.loads in PyJWT._decode_payload (jwt/api_jwt.py:297-300), whose except clause catches only ValueError. A payload that is valid JSON but nested ~20,000 levels deep makes json.loads raise RecursionError, which is not a ValueError and escapes as a raw, …
There is a Re-DoS vulnerability in the is_pem_format function which results in a intesive CPU usage if an attacker is been able to provide a custom certificate.
MemoryMap::read in the pageant crate (part of the russh workspace, used by russh's SSH-agent client on Windows via AgentClient::connect_pageant) copies a peer-controlled number of bytes out of an 8192-byte shared-memory view with no bounds check — unlike the sibling MemoryMap::write, which correctly rejects oversize access with Error::Overflow. The byte count comes straight from a u32 length prefix that the responding "Pageant" process writes into the shared mapping. A malicious local …
nodemailer/lib/addressparser parses one shape of address in O(n^2) time. A single ~640 KB address value blocks the Node.js event loop for roughly 7 seconds. And it is reachable without auth: mailparser feeds inbound email headers straight into this parser, so one crafted email is enough to stall a service that parses mail.
The Node.js ImageResponse implementation from next/og is affected by an upstream vulnerability. This can lead to remote code execution. Affected applications pass attacker-controlled values into SVG content, attributes, or styles during image generation: import { ImageResponse } from 'next/og' export async function GET(request: Request) { const value = new URL(request.url).searchParams.get('value') ?? '' return new ImageResponse( <svg width="1200" height="630"> <title>{value}</title> </svg> ) } Applications using the Edge ImageResponse implementation, or applications …
A peer that can open a TCP connection to a NestJS microservice using the built-in TCP transport can make the server process allocate memory without limit, on either side of the connection, until the process is killed by the OS or by its container memory limit. No authentication, no credentials, and no valid message are required. Applications are affected only if they start a microservice with Transport.TCP and the transport's …
Any authenticated LiteLLM proxy user could redirect an outbound provider call to a destination they control and cause the proxy to send its own configured provider credentials to that destination. The proxy's request-body validation was a denylist that did not cover every sensitive parameter and did not inspect parameters nested inside other request fields, so a caller could supply a routing or credential value that the proxy applied without clearing …
league/commonmark's GitHub Flavored Markdown Table extension registers TableStartParser as a block-start parser. While a paragraph is the active block, the core block parser calls TableStartParser::tryStart() on every non-blank line. Its first action fetches the entire growing paragraph buffer via getParagraphContent() and runs strpos($paragraph, '|') across all of it. For a paragraph of M pipe-free lines, line k rescans about k lines of buffer, so total work is 1+2+…+M, which is …
Summary The DisallowedRawHtml extension does not escape a disallowed tag when the tag name is the last thing in the raw HTML. A Markdown line containing just <script is emitted unchanged, and the next block can supply its attributes. With the shipped GFM defaults this allows stored XSS by anyone who can post Markdown.
With @JsonTypeInfo(use = Id.NAME, defaultImpl = …), every distinct unknown raw type ID selects the same fallback deserializer but is retained as a separate key in TypeDeserializerBase._deserializers. An attacker who can repeatedly supply new unknown type IDs can grow this process-lifetime cache without a configured bound.
With @JsonTypeInfo(use = Id.NAME, defaultImpl = …), every distinct unknown raw type ID selects the same fallback deserializer but is retained as a separate key in TypeDeserializerBase._deserializers. An attacker who can repeatedly supply new unknown type IDs can grow this process-lifetime cache without a configured bound.
When an @JsonIdentityInfo collection or map first creates N unresolved object-ID references and later resolves the same IDs in reverse order, jackson-databind scans the remaining pending-reference accumulator for each resolution. A shallow JSON document whose size grows linearly can therefore cause quadratic CPU work during deserialization.
When an @JsonIdentityInfo collection or map first creates N unresolved object-ID references and later resolves the same IDs in reverse order, jackson-databind scans the remaining pending-reference accumulator for each resolution. A shallow JSON document whose size grows linearly can therefore cause quadratic CPU work during deserialization.
hono/jsx does not HTML-escape a plain string placed directly as a child or fallback of Suspense or ErrorBoundary, as the only child of a Context.Provider, or as the root value of renderToString() / renderToReadableStream() from hono/jsx/dom/server. Such a string is emitted as markup instead of text.
Repo.init decides which directory is the git directory by testing candidate paths in an order that considers the real .git last. Two earlier tests can be satisfied by ordinary tracked files. Git reserves only the literal name .git, so HEAD, objects/, refs/, config, gitdir, commondir and hooks/ at a repository root are all legal tracked content. Consequently, after a victim opens or clones an attacker's repository, GitPython resolves git_dir to …
GitPython's Actor.name_email_regex regular expression (git/util.py, line 863) is vulnerable to catastrophic backtracking (ReDoS — Regular Expression Denial of Service). When GitPython parses the author or committer header of a git commit object that contains a long string with an unterminated < (no matching >), the Python regex engine enters quadratic backtracking, causing complete single-threaded CPU exhaustion proportional to the square of the input length. A single crafted commit object can …
GitPython 3.1.59 blocks a previously available local-file read path through unsafe git diff options such as -O/–orderfile. However, the high-level diff API still permits –no-index with the default allow_unsafe_options=False. –no-index changes the semantics of the paths arguments: instead of repository-relative pathspecs, Git interprets them as arbitrary filesystem paths. When combined with the still-allowed -I/–ignore-matching-lines option, this creates a content-dependent Boolean oracle over a caller-selected local file. This was reproduced against …
Affected: GitPython 3.1.61 (latest release) and main — git/objects/submodule/base.py. git diff 3.1.61 origin/main – git/objects/submodule/ is empty, so both are identical here.
Fastify decided whether to validate a request part by checking its schema for JavaScript truthiness. JSON Schema Draft 7 defines the boolean false as a valid schema that rejects every instance, but because false is falsy, a route that set body, querystring, params, or headers to false had that part left uncompiled: no validator was attached and the request reached the handler. An application that used false as a deny-all …
Fastify runs a route's validator and, for a result shaped like { value, error }, unwraps it: an error becomes a validation failure and value replaces the request part. This convention is intended for synchronous custom compilers (for example Joi). A JSON Schema $async validator, however, resolves with the validated data itself, so Fastify applied the same unwrapping to it. If a request part validated by an $async schema contains …
Fastify lowercases header-schema property names before compiling the schema, because Node.js stores request header names in lowercase. That normalization was incomplete: it lowercased only top-level properties keys and the root required array, and did not lowercase the JSON Schema Draft 7 dependencies keyword (its trigger keys and dependent property names) or names in nested subschemas. As a result, a header schema that uses dependencies to require one header when another …
fastify crashes with an uncaught ERR_HTTP2_INVALID_CONNECTION_HEADERS exception when a route that registers a response trailer via reply.trailer() is served over HTTP/2. Fastify unconditionally adds the Transfer-Encoding: chunked header when a trailer is set, which is forbidden on HTTP/2, so Node.js throws while serializing the response headers. The exception is not caught and becomes an uncaughtException, terminating the Node.js process. One unauthenticated HTTP/2 request to any route that uses trailers is …
Fastify routes a malformed URL under one plugin prefix to the custom not-found handler of a different sibling plugin, invoking the handler registered last and skipping the preHandler declared in its setNotFoundHandler(). When the request method has no route in the main router, a malformed request target reaches Fastify's internal not-found router before URL decoding and is dispatched through a single shared handler pointer, regardless of prefix and without the …
In IN_PLACE mode DOMPurify sanitizes the caller's live DOM subtree directly. To close a known hazard (GHSA-55q2-fjhq-7xh7), the library neutralizes any node a hook detaches during sanitization by stripping its subtree's non-allow-listed attributes, but this neutralization is wired only into the beforeSanitizeElements and uponSanitizeElement hook sites. An afterSanitizeElements or afterSanitizeAttributes hook that removes a non-root element (a documented, supported pattern) detaches that element's subtree with no neutralization, so descendant on* …
Axios for Node.js parses data: URLs in lib/helpers/fromDataURI.js. The current RFC-2397 parser uses a regular expression whose media type groups allow / inside both sides of the type/subtype match. A malformed data: URL containing many slashes and no comma forces the JavaScript regex engine to try many possible placements for the separator before failing. Applications are affected when they pass untrusted URL strings to axios and do not reject or …
Axios shouldBypassProxy() normalizes hostnames with hostname.replace(/.+$/, ''). For a hostname shaped as many dots followed by a non-dot, the anchored regex can perform quadratic backtracking. Because axios re-evaluates proxy bypass rules for redirected requests, a malicious server can trigger this synchronous work through a crafted redirect Location. The issue affects Node.js applications that use environment proxy variables with NO_PROXY and allow redirects. shouldBypassProxy normalizes a host with hostname.replace(/.+$/, ''). On …
Axios default-instance requests that omit an explicit method can read an inherited method value from Object.prototype. If another vulnerability in the same process pollutes Object.prototype.method, calls such as axios.request({ url }) and axios({ url }) can send a state-changing HTTP method instead of the expected default GET. Axios does not create the prototype pollution source. This is a read-side gadget in axios request dispatch. Axios 1.17.0 contains a read-side prototype-pollution …
Axios fetch adapter requests can be altered by inherited properties on fetchOptions. The adapter resolves method, headers, body, signal, and credentials into resolvedOptions, creates a Request, and then calls fetch(request, fetchOptions) instead of fetch(request, resolvedOptions). In runtimes such as Node's undici-backed fetch, inherited fetchOptions.headers can override the headers already placed on the Request. Axios does not create the prototype pollution source. This is a read-side gadget that becomes exploitable after …
Axios form serialization reads visitor, maxDepth, dots, indexes, metaTokens, and Blob from an internal options object without own-property guards. When Object.prototype has been polluted elsewhere in the same process, those inherited values can change how axios serializes multipart and URL-encoded request bodies. Axios does not create the prototype pollution source. This is a read-side gadget: axios turns an existing same-process pollution condition into altered request serialization or request failures. axios …
Axios' Node HTTP adapter can act as a read-side prototype-pollution gadget for Node's sensitive createConnection request option. The adapter creates a null-prototype options object, but Node's HTTP client can copy or normalize request options into ordinary objects before connection creation. If Object.prototype.createConnection has been polluted elsewhere in the same process, Node can call the inherited function and create a socket to an attacker-controlled endpoint. Axios does not create the prototype …
Axios exposes maxRedirects to limit redirect following, and maxRedirects: 0 is used by applications as a redirect-based SSRF guard. The Node HTTP adapter enforces this option. The fetch adapter does not read it and does not set a Fetch API redirect mode, so the runtime default of redirect: 'follow' applies. Applications are affected when they rely on maxRedirects: 0 and use the fetch adapter, either explicitly or because the runtime …
Axios for Node.js does not apply configured DNS lookup or proxy controls when a request uses httpVersion: 2. The HTTP/1 adapter path wraps and forwards config.lookup, builds normal request options, and applies proxy routing through setProxy(). The HTTP/2 path builds a session with http2.connect() using only options.http2Options, which drops the top-level lookup, agent, and proxy state. Applications are affected when they allow a user to influence request destinations, enable axios …
Axios request interceptors may return a replacement config object. If an interceptor returns a plain object without an own headers property, dispatchRequest() later evaluates config.headers and can resolve an inherited Object.prototype.headers value. In a process where another vulnerability has polluted Object.prototype.headers, axios can send attacker-controlled headers. Axios does not create the prototype pollution source, and the interceptor itself is trusted caller code. The vulnerable behavior is the post-interceptor axios config …
Axios contains a guard in the Node HTTP adapter to avoid using an inherited Object.prototype.getHeaders as a FormData header source. The fetch adapter calls the shared resolveConfig() helper before dispatch, and that helper lacks the same guard. If another vulnerability pollutes Object.prototype with FormData-like properties and getHeaders(), the fetch adapter can merge attacker-controlled headers into the outbound request. Axios does not create the prototype pollution source. This is a read-side …
Axios versions with Node.js HTTP/2 support can terminate the caller’s process when a ClientHttp2Session emits an error event that is not handled by axios. This affects applications that use the Node HTTP adapter with httpVersion: 2. A malicious, unavailable, or non-HTTP/2 endpoint can cause an uncaught exception instead of a normal rejected axios request.
Axios supports proxy environment variables and evaluates NO_PROXY exclusions in the Node.js adapter. CIDR-form NO_PROXY entries such as 127.0.0.0/8, 10.0.0.0/8, or 169.254.169.254/32 are not interpreted as IP ranges. As a result, a request to an IP address inside a configured CIDR exclusion can still be sent through the configured proxy. This affects deployments that rely on CIDR notation to keep loopback, private, Kubernetes, CI, or cloud metadata traffic away from …
The @astrojs/netlify adapter generated regular expressions for Netlify Image CDN remote-image allowlists without anchoring them to the beginning of the URL. Netlify evaluates these expressions with RegExp.test(), so an allowed image origin appearing anywhere in a URL, including its path or query string, could satisfy the allowlist. For example, an application allowing images.example.com could accept a source URL such as: http://127.0.0.1:6379/?url=https://images.example.com/image.png The allowlist matched the approved origin in the query …
In the Astro Node adapter, a request whose Host header contains a malformed port (for example example.com:65536 or example.com:8080:8080) produced an invalid request URL. The fallback intended to recover from an unparseable URL reused the same malformed host, so it failed again and raised an uncaught TypeError: Invalid URL while the request was being built, before any route ran.
A denial of service (DoS) vulnerability was identified in @angular/router when Server-Side Rendering (SSR) is enabled on Node.js (V8). When @angular/router parses incoming request URLs, it extracts path segments, matrix parameters, and child outlets into plain JavaScript objects (Record<string, string>). When matrix parameter names or outlet names are numeric strings (such as /a;990;2522), the V8 JavaScript engine interprets them as array-indexed properties rather than named properties. Under V8's internal property-storage …
On the Fastify adapter, an HTTP request that uses an absolute-form request target (GET http://host/path HTTP/1.1 instead of GET /path HTTP/1.1) reaches the route handler without running the path-scoped Nest middleware bound to that route. Applications that enforce authentication or authorization in middleware execute the protected handler with those checks skipped.
If an application method handler crashes, the error message is included in the status message sent to the client. This can leak to the client any sensitive data that may be included in the error message. This impacts anyone using @grpc/grpc-js to run servers.
When server credentials are created with the requireClientCertificate option set to false, getAuthContext does not distinguish between authorized and unauthorized certificates in its return value. This can create improper authentication vulnerabilities for @grpc/grpc-js users who use the result of getAuthContext for authentication. In particular, @grpc/grpc-js-xds can both set the requireClientCertificate option to false and use the return value of getAuthContext for RBAC authentication in some configurations.
[!IMPORTANT] CVE-2026-76844 was assigned and published for this issue by VulnCheck on 2026-08-24 without prior coordination with the webpack maintainers or the OpenJS Foundation, which holds the CVE Numbering Authority scope for webpack projects. Neither the maintainers nor the OpenJS CNA were notified before publication, and no fix was available at that time. This advisory is the coordinated record produced by the webpack maintainers and the OpenJS Foundation CNA.
undici's BalancedPool passes its constructor options through a JSON-based deep clone (JSON.parse(JSON.stringify(…))) before forwarding them to each per-upstream Pool. JSON cannot represent functions, so a caller-supplied connect or tls option containing a checkServerIdentity callback (or a custom connector function) is silently dropped before it reaches the TLS layer. As a result, a TLS peer whose certificate a custom checkServerIdentity was written to reject, but which passes Node's default hostname and …
undici's interceptors.dump() reads and discards response bodies up to a configurable maxSize. When a response declares a Content-Length that exceeds maxSize, the request is aborted cleanly. When a response is sent chunked (no Content-Length) and its body exceeds maxSize, it is not aborted: the interceptor ends the response early once the accumulated size reaches maxSize, and continued delivery from the parser triggers an internal assertion that is caught and turned …
Undici's interceptors.retry() can resume a request after a partial response and append the resumed bytes to an already partially delivered body, while the application still receives the original response's status and headers. When that response carried a Content-Length, the application can receive a longer body. Applications that forward Undici's status, headers, and body downstream without recalculating framing, for example proxy or gateway applications, may emit a response whose body exceeds …
undici's WebSocketStream crashes the client process when a WebSocket connection is closed abruptly without a close handshake. On such an unclean close, the internal socket-close handler calls abort() on the writable stream even when the application holds a writer lock. Per the WHATWG Streams standard, aborting a locked stream returns a promise that rejects with a TypeError, and the handler discards that promise. The unobserved rejection surfaces as an unhandledRejection …
The undici WebSocket client throws an uncaught TypeError during the opening handshake when a server's 101 response includes a Sec-WebSocket-Protocol header that the client never requested. The throw occurs in a queueMicrotask callback with no surrounding try/catch, so it propagates as an uncaught exception and terminates the Node.js process. This is a remote, unauthenticated denial of service against any application that opens a WebSocket to an attacker controlled or compromised …
The interceptors.decompress() interceptor decompresses HTTP response bodies according to the untrusted Content-Encoding header. The number of decompression layers is capped at 5, but the total decompressed output size is not bounded and there is no option to limit it. A malicious or faulty upstream can return a small compressed payload (a compression bomb) that expands to hundreds of megabytes or gigabytes in client memory, exhausting memory and causing the Node.js …
undici's RetryHandler can leave a response body pending indefinitely. When a retried request receives a non-retryable response after a truncated one, the original response.body held by the application is never settled, so reads such as response.body.text() hang and bodyTimeout does not fire. A malicious server can repeat this to accumulate pending promises and streams, leading to denial of service.
undici's interceptors.cache() does not handle Set-Cookie in the cache path. In shared-cache mode (type: 'shared', the default), a cacheable response (for example Cache-Control: public, max-age=…) carrying a Set-Cookie header is stored, and the stored Set-Cookie is re-served to a later caller that hits the same cache key. This exposes one user's cookie to another caller and lets an untrusted upstream inject cookies into cached responses served to all subsequent callers, …
When interceptors.cache() or interceptors.deduplicate() is used with a dispatcher that does not carry a single authoritative origin, or when a request supplies its own origin, undici builds the cache and deduplication keys without the actual destination origin. If a cache store or interceptor instance is shared across more than one origin, otherwise-identical requests to different origins are keyed together. An attacker who controls the response from one origin can then …
undici's interceptors.cache() documents that it caches only safe HTTP methods. However, its internal skip-list is built by subtracting the configured methods from the safe-methods set, so an unsafe method (POST, PUT, PATCH, DELETE) never lands in the skip-list and is looked up against the cache store. Combined with the storage gate (canCacheResponse) having no method check, a heuristically-cacheable response (for example a 404) with an explicit Cache-Control: max-age=… to an …
A denial-of-service vulnerability exists in Engine.IO / Socket.IO servers that allow transport upgrades. The Engine.IO protocol revision is negotiated during the initial handshake and stored on the session, but a newly-created transport, including a WebSocket upgrade transport, could independently derive a different protocol revision from the upgrade request query parameters. The server did not verify that the protocol revision of an upgrade request matched the protocol revision of the existing …
PyJWS._looad() (jwt/api_jws.py:337) splits the compact token, base64url decodes the header segment and hands it to json.loads() before _verify_signature() runs. The parse is wrapped in except ValueError as e: raise DecodeError(…). That is what makes the documented contract sufficient for untrusted input: except jwt.PyJWTError (or InvalidTokenError) around jwt.decode(token, key, algorithms=[…]). CPython's JSON decoder raises RecursionError (a RuntimeError subclass, not a ValueError) once nesting crosses the native stack guard. That exception leaves …
Summary CVE-2026-48524 (GHSA-fhv5-28vv-h8m8, "PyJWKClient unbounded JWKS endpoint requests via attacker-controlled kid values (DoS)") was fixed in 2.13.0 by stopping fetch_data() from clearing the cache on a fetch error. That closed one amplification path but did not add the mitigation the advisory's title implies: there is still no rate-limit, negative-cache, or minimum-refresh-interval for unknown kids. At HEAD, get_signing_key(kid) (jwt/jwks_client.py:185-211), on any unknown kid, calls get_signing_keys(refresh=True), and refresh=True bypasses jwk_set_cache unconditionally and …
PyJWT 2.13.0 PyJWKClient followed HTTP redirects while fetching a JWKS, without validating the redirect destination. A configured trusted endpoint could therefore redirect the client to a different host.
A service that verifies HS256 tokens using an empty oct JWK through PyJWK, including a PyJWK obtained from PyJWKSet, can therefore accept attacker-generated tokens as authenticated. PyJWT 2.13.0 rejects an empty HMAC key when it is supplied through the raw str/bytes key path, but accepts the same zero-length key when it is supplied as a symmetric PyJWK. An oct JWK containing an empty Base64URL key value: {"kty":"oct","k":""} is decoded to …
HMACAlgorithm.prepare_key blocks asymmetric keys from being used as HMAC secrets by searching for text markers only. It looks for —–BEGIN and for an ssh- prefix. The same key in DER form is binary ASN.1 and has neither marker, so it passes the check and is used as an HMAC secret. An application that verifies tokens with an RSA or EC public key, and also allows HS256 with that same key, …
PyJWT 2.13.0 accepts compact JWS signature segments containing characters that are not in the Base64URL alphabet. Appending !!!! to a valid signature does not change the decoded signature bytes or authenticated claims, but it changes the serialized token and its SHA-256 hash. An application that indexes logout or revocation state by the raw token therefore rejects the logged-out token and accepts its no-key alternate serialization. This report separates two issues: …
A malformed RSA JWK inside a JWK Set aborts parsing of the entire set instead of being skipped, because RSAAlgorithm.from_jwk can raise a plain ValueError that isn't caught by PyJWKSet's per-key error-skipping logic.
A PEM public key with marker-adjacent whitespace, CR-only line terminators, or folded to a single line makes PyJWT's is_pem_format() return False while cryptography.load_pem_public_key() accepts the identical bytes. The asymmetric-key rejection in HMACAlgorithm.prepare_key is skipped, the public key becomes the HMAC secret, and an attacker who knows the public key mints a valid HS256 token — universal forgery — whenever the verify allow-list mixes an HMAC and an asymmetric algorithm.
An unauthenticated network attacker who knows the target application's RSA public key — which is public by design and obtainable from a JWKS endpoint or certificate — can forge JWT tokens containing arbitrary claims and have them accepted by a PyJWT 2.13.0 verifier configured with a mixed algorithm set (algorithms=["HS256", "RS256"] or equivalent). The resulting impact is complete authentication and authorization bypass (C:H/I:H). Attack Complexity is High (AC:H) because the …
PyJWT 2.13.0 contains an incomplete defense against algorithm confusion when an application mixes symmetric and asymmetric algorithms in one verification path. A public RSA, EC, or OKP JWK can be accepted as an HMAC secret when it is wrapped in a JWKS object, nested in an array, or represented in another container form that does not expose a top-level kty member.
PHPCSUtils versions 1.0.0-alpha1 through 1.2.2 contain an arbitrary code execution vulnerability in PHPCSUtils\AbstractSniffs\AbstractArrayDeclarationSniff::getActualArrayKey(). The vulnerable method is reached by any sniff that extends AbstractArrayDeclarationSniff and calls getActualArrayKey(). Running PHPCS over untrusted PHP code through such a sniff, for example, in a CI pipeline that lints pull requests, or on a developer machine reviewing third-party code, could lead to arbitrary command execution on the scanning host. The vulnerability happens when the …
The OpenTelemetry Go SDK trace package can fail to enforce AttributeValueLengthLimit for string attributes containing the valid Unicode replacement character U+FFFD. An oversized attacker-controlled attribute value that includes U+FFFD is returned untruncated, bypassing the configured memory/DoS protection and allowing increased per-span memory usage. The finding is low severity because it requires a deployment with attribute value length limits enabled and attacker-controlled data being recorded into span attributes. Introduced in commit …
A BatchingProcessor in go.opentelemetry.io/otel/sdk/log can enter a tight CPU loop when the asynchronous export buffer is full. Under exporter backpressure, attacker-driven high-volume log emission can keep the queue at or above the batch size, causing repeated immediate export retries and a denial of service through CPU exhaustion. Introduced in commit: 4af9c20
A timing side-channel vulnerability exists in the PKCE (RFC 7636) implementation of the Authorization Code Grant flow. The code_challenge_method_plain function uses Python's standard == operator for string comparison instead of a constant-time comparison function, potentially allowing timing-based attacks.
When enable_jsonp=True, oauthlib's RevocationEndpoint reflects the user-supplied callback parameter directly into JavaScript response bodies on both success and error paths without validating that it is a legal JSONP callback name. This allows arbitrary JavaScript response generation instead of a restricted function call, making the documented JSONP revocation feature unsafe for browser-based JSONP consumption when attackers can influence callback.
Nodemailer's address parser can produce an unexpected recipient address when an RFC 5322 comment follows the domain of an address whose local-part is a quoted string. For example: "user"@example.com(x)evil.com is parsed as: { address: "user@example.com evil.com", name: "" } The resulting address therefore contains additional attacker-controlled domain text separated by a literal space. The parsed .address value is subsequently propagated into the SMTP envelope: src/addressparser/index.ts ↓ recipient.address ↓ src/mime-node/index.ts ↓ …
When addressparser finds no address by its strict reading, it falls back to pulling one out of the free text with /\s*\b[^@\s]+@[^\s]+\b\s*/. That pattern backtracks quadratically: [^@\s]+ is retried from every offset and rescans the run to the next @ each time. A single header value holding a long whitespace-free run with no usable @ blocks the Node.js event loop for tens of seconds. It is cheaper to exploit than …
Nodemailer 10.0.1 does not safely process deeply nested arrays supplied through recipient fields such as to, cc, and bcc. The public MimeNodeAddressInput type recursively permits arrays, but MimeNode._parseAddresses() flattens only the outermost array. A remaining nested array is passed to addressparser(), whose Tokenizer coerces the input with .toString(). Native Array.prototype.toString() recursively processes every nested array through join() and toString() until the V8 call stack is exhausted. A valid 10,021-byte JSON …
A single message whose pattern is a deeply nested object terminates a NestJS microservice that uses the TCP or RabbitMQ transport. The server serialized the client-supplied pattern with JSON.stringify to derive the handler lookup key; on deeply nested input this throws RangeError: Maximum call stack size exceeded. The exception escaped the asynchronous message handler as an unhandled promise rejection, which terminates the Node.js process under the default –unhandled-rejections=throw.
moment before 2.31.0 is vulnerable to path traversal in moment.locale(). When an application passes a non-string, attacker-influenced value to moment.locale(), a specially crafted object can bypass the locale name validation and cause moment to load a file from an attacker-controlled path. This is a further bypass of the validation added in 2.29.2 for CVE-2022-24785. This affects server-side (npm) users only. Plain string input is not affected: the existing validation correctly …
Two independent quadratic paths in the linkify: true handling. Both are in markdown-it's own code rather than in linkify-it, which stays linear on both payloads. src/rules_core/linkify.ts calls arrayReplaceAt once per linkified text token, and that rebuilds the whole children array each time. A paragraph of N soft-broken lines is one inline token with about 2N children, so you get N rebuilds over a 2N array. Schema-less emails are what reach …
When APP_DEBUG=true, attacker-controlled input is passed to a Tippy.js tooltip configured with allowHTML: true, enabling DOM-based XSS during mouse hover.
maxTotalMergeKeys does not count empty mappings. An attacker can repeatedly merge a large sequence of them and consume significant CPU without reaching the configured limit.
Any application that validates a user-supplied string with Joi.string().isoDate() can be stalled by a single request. One of the regular expressions the rule runs over the input was unanchored, so a valid ISO date followed by a long run of fractional-second digits made the regex engine restart its search from every position in the string, costing time proportional to the square of the input length. 64 KB of digits costs …
isInSubnet() and isHostInSubnet() accept an address of either family and compare masked binary strings without checking that both operands are the same family. Address4 pads to 32 bits and Address6 to 128, so whenever the leading bits agree the strings are equal: new Address6('a00::1').isInSubnet(new Address4('10.0.0.0/8')) is true, and new Address4('32.0.0.1').isInSubnet(new Address6('2000::/3')) is true. No IPv4 address is inside an IPv6 network, so both answers are untrue. An application that parses …
new Address6() and Address6.isValid() place no bound on the length of the string they parse. When the string contains a character that cannot appear in an IPv6 address, the parser builds a diagnostic that wraps every such character in a 34-byte <span class="parse-error">, so the work and the memory scale with the input rather than with an address. A 1 MiB string of ! costs about 110 MB and 70 …
OS Command Injection in internal/gitcmd (git diff filter.clean/smudge invocation) in esengine DeepSeek-Reasonix (Reasonix Studio) allows a local attacker who controls repository content (.gitattributes + .git/config) to execute arbitrary commands via the desktop app's workspace-changes diff viewer.
Related public issue (context, not a duplicate) Closed issue #1429 ("Handle non-UTF-8 paths", 2024-03-24) raised exactly this general concern and even suggested detection via preg_match('//u', $path) !== 1 – note the reporter's suggested check explicitly compares !== 1, which would correctly treat PCRE's false return as "reject." The maintainer's reply pointed to the PathNormalizer interface as the place to implement this. The control-character check that ended up shipping in WhitespacePathNormalizer …
fast-uri's mailto scheme parser compares each query field name to the reserved names (to, subject, body) while the name is still percent-encoded, and only percent-decodes it when storing it as a generic header. On serialize, the decoded name is re-emitted, so a field name such as %74o (percent-encoded to) is not recognized as a recipient at parse time (parse().to shows only the legitimate recipient) but materializes as a literal to= …
fast-uri folds the host to lowercase before it percent-decodes the host, so a percent-encoded uppercase unreserved octet such as %41 decodes to a literal A that is never folded. For a scheme-relative reference (//host) there is no scheme, so the host canonicalization that repairs this on a scheme-bearing URL does not run. As a result parse, normalize, and equal disagree on the same host: parse("//%41.com").host returns "A.com" while parse("//a.com").host and …
Windows opened from a sandboxed top-level document did not inherit that document's HTML sandbox restrictions, so content that was meant to run sandboxed could open a window with the app's full origin. GHSA-hq2x-r82h-9wj4 covers the same issue for sandboxed iframes. Apps are only affected if they render untrusted content in a sandboxed top-level document that allows popups. Apps that deny popups from untrusted content with setWindowOpenHandler are not affected.
The cache Electron keeps for sandboxed preload scripts did not verify that a cached entry matched the preload it was served for. A compromised renderer could use this to run its own code in the preload context on a later load. Apps are only affected if they load untrusted content. Apps that do not load untrusted content are not affected.
On macOS, Electron bundles the Squirrel.Mac auto-update framework, whose privileged ShipIt helper performs the final step of an update as root. A local attacker could cause that helper to overwrite a different application's files, as root, instead of the app that started the update. Exploitation requires local access to the machine. Apps are only affected on macOS if they ship Squirrel.Mac-based auto-updates. Apps on other platforms, or that do not …
Responses served through protocol.registerFileProtocol or protocol.registerHttpProtocol for a custom scheme registered with supportFetchAPI: true but without corsEnabled: true could be read cross-origin by web content. This completes the fix for CVE-2026-70604. Apps are only affected if they register such a scheme, serve it through one of those handlers, and load untrusted content. Apps that set corsEnabled: true, or that do not load untrusted content, are not affected.
A <webview> could enable Node.js integration in its Web Workers even when its embedder had Node.js integration disabled, giving guest content more privilege than the embedder allowed. Apps are only affected if they enable the <webview> tag and the embedder is unsandboxed. Apps that do not use <webview>, or that keep the embedder sandboxed, are not affected.
Popups opened from a sandboxed iframe through a link (for example target="_blank" or a middle-click) did not inherit the iframe's HTML sandbox restrictions. Content that was meant to run sandboxed could open a popup with the embedding app's full origin, gaining access to that origin's cookies, storage, and same-origin scripting. Apps are only affected if they embed untrusted content in iframes sandboxed with allow-scripts allow-popups. Apps that do not embed …
Expanding {a},b}-shaped input takes time quadratic in the number of literal } characters, blocking the event loop. Bash preserves a quirk where a brace group followed by a comma set still expands ({a},b}). The parser implements this by rewriting the string and restarting the scan. Each pass absorbs exactly one } and re-scans from the beginning, so n trailing braces cost n full passes.
expand_() recurses once per level of brace nesting. Deeply nested input exhausts the native stack and crashes the process. This is distinct from CVE-2026-14257 / GHSA-mh99-v99m-4gvg, which made the tail iterative (recursion on m.post, driven by how many groups are chained). Nesting depth drives a different recursion that the tail fix never touched, so the documented constant-stack-depth guarantee only ever covered chained input, not nested input. It is also distinct …
parseCommaParts() can exhaust the native stack and crash the process. There are two distinct ways to trigger it, both reachable from a single untrusted pattern string. This is the parsing-side counterpart to CVE-2026-14257 / GHSA-mh99-v99m-4gvg. That fix made expand_() iterative and documented a constant-stack-depth guarantee, but parseCommaParts() was left recursive, so the guarantee only held for one of the two parsing paths.
The following SVG will produce a link with a javascript scheme. If the user clicks this link, they will run it. <svg xmlns="http://www.w3.org/2000/svg"> <a> <set attributeName="href" to="javascript:alert('SET_XSS')"></set> <text y="30">Click set</text> </a> </svg>
Denial of Service in adm-zip's async decompression API allows an unauthenticated attacker to crash the entire Node.js host process by supplying a single malformed ZIP file.
adm-zip enforces a maxOutputLength guard against decompression bombs on its synchronous getData() path, but the equivalent asynchronous getDataAsync() path does not enforce it — it accumulates and returns the entire decompressed output regardless of the entry's declared size. An application that checks an entry's declared size before deciding to process it, then reads the entry via getDataAsync() (a completely ordinary, often-recommended choice for I/O in Node.js), gets none of the …
A ZIP file can contain two entries with the identical name. adm-zip keeps both in its internal entry list, but its name-lookup table only retains the last one written. getEntry(name) and extractAllTo() walk these two different internal structures, so they can each resolve a duplicate name to a different entry. An application that validates a named entry's contents via getEntry() before trusting an archive, then extracts the whole archive, can …
The fix shipped for CVE-2026-39244 (methods/inflater.js) caps zlib's decompression output via maxOutputLength: expectedLength, where expectedLength is read directly from the ZIP entry's attacker-controlled "uncompressed size" header field (CENLEN/LOCLEN). This cap is only applied when expectedLength > 0: const option = version >= 15 && expectedLength > 0 ? { maxOutputLength: expectedLength } : {}; return zlib.inflateRawSync(inbuf, option); If an attacker sets the declared uncompressed-size field to exactly 0, this condition …
adm-zip applies the Unix permission bits stored in a zip entry directly to the extracted file via fs.chmodSync() when keepOriginalPermission=true is passed to extractAllTo()/extractEntryTo() — and it never filters the setuid/setgid/sticky bits out of those bits. A zip crafted by an attacker can therefore produce an extracted binary with mode 04755. When extraction runs as root (the default posture in Docker builds, CI runners, and privileged install steps — the …
When extracting an untrusted archive with the default decompress(input, output) API, a crafted archive containing a chain of symlink entries can make a later entry resolve outside the output directory. The lexical containment checks pass, but the kernel follows the planted symlinks to a path outside output, letting an attacker write (and read) files outside the intended extraction directory. Overwriting startup scripts or configuration can lead to remote code execution. …
When extracting an untrusted archive with the default decompress(input, output) API, a crafted archive containing a chain of symlink entries can make a later entry resolve outside the output directory. The lexical containment checks pass, but the kernel follows the planted symlinks to a path outside output, letting an attacker write (and read) files outside the intended extraction directory. Overwriting startup scripts or configuration can lead to remote code execution. …
undici's WebSocket client (including Node.js's bundled globalThis.WebSocket) crashes the entire Node.js process when a remote WebSocket peer sends a permessage-deflate compressed message that crosses the decompressed-payload size limit and then contains a malformed DEFLATE block. In lib/web/websocket/permessage-deflate.js, the size-limit cleanup calls removeAllListeners() on the internal zlib InflateRaw, removing its error listener, but leaves the stream running. The inflater then emits a Z_DATA_ERROR with no listener attached, which Node.js treats as …
Incomplete Prototype Pollution Fix Allows Mutation of Inherited Built-in Method Objects scim-patch blocks direct dangerous path segments such as proto, constructor, and prototype, but still traverses inherited properties when applying SCIM patch paths. An attacker who controls a SCIM PATCH operation can use paths such as toString.polluted to mutate shared built-in function objects, for example Object.prototype.toString.
Nodemailer's process-global DNS cache is keyed only by host, but each cache entry also stores the caller-specific TLS servername. When two direct SMTPS transports use the same DNS host with different tls.servername values, the first transport's server name is returned to the second transport and overwrites its explicitly configured value. As a result, Nodemailer sends the wrong SNI value and verifies the peer certificate against the wrong identity. In a …
Multer's diskStorage can leave complete, orphaned files on disk when a multipart upload is aborted in the brief window before the storage engine assigns the file path. This is an incomplete fix of CVE-2026-5038: the earlier cleanup removes only in-flight uploads that already have a path, so an upload aborted inside that window is written to disk with nothing left to remove it, and the application is not given a …
Morgan writes attacker-controlled request data to the access log. Its escaping (added in 1.11.0 and 1.12.0) neutralizes control characters, the Unicode line separators, and backslash, but not the double quote (0x22), which is the field delimiter of the Apache combined log format morgan emits. An attacker who controls a quoted field (:user-agent, :referrer, the request URL, or the Basic auth :remote-user) can inject a double quote to close the field …
A java.nio.file.Path field bound from untrusted JSON reaches JDKFromStringDeserializer.NioPathHelper.deserialize. The attacker string flows through new URI(value) → Path.of(uri), then on FileSystemNotFoundException into a ServiceLoader<FileSystemProvider> enumeration that calls provider.getPath(uri) on the first scheme-matching provider. No scheme is rejected, so untrusted JSON can drive an arbitrary registered provider under the default JsonMapper.builder().build(). Impact is bounded. The JDK built-in providers (file, jar/zipfs) do no network I/O and do not mount, so the path …
A java.nio.file.Path field bound from untrusted JSON reaches JDKFromStringDeserializer.NioPathHelper.deserialize. The attacker string flows through new URI(value) → Path.of(uri), then on FileSystemNotFoundException into a ServiceLoader<FileSystemProvider> enumeration that calls provider.getPath(uri) on the first scheme-matching provider. No scheme is rejected, so untrusted JSON can drive an arbitrary registered provider under the default JsonMapper.builder().build(). Impact is bounded. The JDK built-in providers (file, jar/zipfs) do no network I/O and do not mount, so the path …
CVE-2026-54514 (GHSA-hgj6-7826-r7m5) fixed an eager-DNS-resolution / SSRF issue in jackson-databind's deserialization of java.net.InetSocketAddress by switching to InetSocketAddress.createUnresolved(…) (PR #5951, commit 1f5a1037, released in 2.18.8 / 2.21.4 / 3.1.4). That fix did not cover the sibling java.net.InetAddress branch in the very same FromStringDeserializer.Std._deserialize() switch statement, which still calls InetAddress.getByName(value) and therefore performs an eager forward DNS lookup on attacker-controlled input at deserialization time. The fix is incomplete: the same vulnerability class …
CVE-2026-54514 (GHSA-hgj6-7826-r7m5) fixed an eager-DNS-resolution / SSRF issue in jackson-databind's deserialization of java.net.InetSocketAddress by switching to InetSocketAddress.createUnresolved(…) (PR #5951, commit 1f5a1037, released in 2.18.8 / 2.21.4 / 3.1.4). That fix did not cover the sibling java.net.InetAddress branch in the very same FromStringDeserializer.Std._deserialize() switch statement, which still calls InetAddress.getByName(value) and therefore performs an eager forward DNS lookup on attacker-controlled input at deserialization time. The fix is incomplete: the same vulnerability class …
jackson-databind 3.2.1 deserializes a JSON string bound to a javax.xml.datatype.Duration or javax.xml.datatype.XMLGregorianCalendar field by passing the raw string verbatim to DatatypeFactory.newDuration(value) / newXMLGregorianCalendar(value). Per the XML-Schema lexical grammar these factory methods accept numeric components of arbitrary length, which the JDK materializes into java.math.BigInteger / BigDecimal using the native BigInteger(String) constructor (an O(n²) parser). Because the digits reside inside a JSON string token, jackson-core's StreamReadConstraints.maxNumberLength guard (which bounds only JSON number …
jackson-databind 3.2.1 deserializes a JSON string bound to a javax.xml.datatype.Duration or javax.xml.datatype.XMLGregorianCalendar field by passing the raw string verbatim to DatatypeFactory.newDuration(value) / newXMLGregorianCalendar(value). Per the XML-Schema lexical grammar these factory methods accept numeric components of arbitrary length, which the JDK materializes into java.math.BigInteger / BigDecimal using the native BigInteger(String) constructor (an O(n²) parser). Because the digits reside inside a JSON string token, jackson-core's StreamReadConstraints.maxNumberLength guard (which bounds only JSON number …
DefaultBaseTypeLimitingValidator — the PolymorphicTypeValidator used automatically whenever @JsonTypeInfo is applied without an explicitly configured custom validator — denies polymorphic resolution only for nine specific "unsafe base types" (Object, Serializable, Closeable, AutoCloseable, Cloneable, Runnable, java.util.logging.Handler, javax.naming.Referenceable, javax.sql.DataSource). Its isSafeSubType() returns true unconditionally for every other base type. java.lang.Comparable is not in that list, despite being implemented by a very large fraction of JDK and application classes — comparable in breadth to …
DefaultBaseTypeLimitingValidator — the PolymorphicTypeValidator used automatically whenever @JsonTypeInfo is applied without an explicitly configured custom validator — denies polymorphic resolution only for nine specific "unsafe base types" (Object, Serializable, Closeable, AutoCloseable, Cloneable, Runnable, java.util.logging.Handler, javax.naming.Referenceable, javax.sql.DataSource). Its isSafeSubType() returns true unconditionally for every other base type. java.lang.Comparable is not in that list, despite being implemented by a very large fraction of JDK and application classes — comparable in breadth to …
No classifier on Address6 recognizes the NAT64 local-use range 64:ff9b:1::/48 (RFC 8215). isPrivate(), isLoopback(), isLinkLocal() and their siblings all return false for every address in it, so an internal IPv4 destination written through a local-use NAT64 prefix (64:ff9b:1:7f00:0:100:: for 127.0.0.1, 64:ff9b:1:a9fe:a9:fe00:: for 169.254.169.254) reads as an ordinary global address. getType() names the range 'NAT64 (local-use)', so the library knows what the address is and classifies it as nothing. An application …
Address6.isLinkLocal() recognizes fe80::/64 rather than fe80::/10. Link-local unicast is the whole /10 under RFC 4291 §2.4 and the IANA IPv6 Special-Purpose Address Registry, so the method returns false for every link-local address outside the one /64 that stateless address autoconfiguration happens to use. new Address6('fe81::1').isLinkLocal() is false. The library contradicts itself on the same object: for fe81::1, getType() returns 'Link-local unicast', getScope() returns 'Link local', and isHostInSubnet(new Address6('fe80::/10')) returns true, …
fast-uri accepts a host that contains an unbalanced or misplaced authority bracket ([ or ]) without reporting an error. A host that starts with [ but does not end with ], such as [@127.0.0.1, is neither validated as an IP literal nor canonicalized as a domain name, so parse() returns it as the host with error undefined, while Node's URL (and http.get, axios, got, and other clients built on it) …
fast-uri serializes the port component of a URI without validating it. When recomposing the authority, fast-uri escapes the userinfo and host components but concatenates the port verbatim, so a port value that is not a sequence of digits can inject authority delimiters. For example, serializing a component whose port is @127.0.0.1:8124 produces http://trusted.example:@127.0.0.1:8124/app, demoting the intended host to userinfo and pointing the authority at an attacker-controlled host. Both fast-uri and …
The grep_search tool in code-ollama constructs a shell command string by interpolating attacker-controlled pattern and path arguments, then executes it via child_process.exec(). The sanitization only escapes backslashes and double-quote characters, leaving $() command substitution and backtick expansion fully intact. A malicious or compromised Ollama server can therefore inject and execute arbitrary OS commands with the privileges of the local user running code-ollama. Because grep_search is classified as a read-only tool, …
An XSS vulnerability exists in @angular/platform-server during server-side rendering (SSR) HTML serialization of ProcessingInstruction DOM nodes (<?target data?>, nodeType === 7) when nested inside fallback raw-content elements (<noscript>, <iframe>, <noembed>, <noframes>). While processing instruction data escaped > to >, it did not check for or escape matching closing tags of ancestor fallback elements (e.g., </noscript>). When rendered in a browser with scripting enabled, an unescaped closing tag sequence in a …
A Denial of Service (DoS) vulnerability exists in @angular/platform-server's DOM emulation parser (domino). When processing untrusted user input containing an incomplete DOCTYPE declaration ending with whitespace before EOF (such as <!DOCTYPE html ), the HTML parser enters an infinite synchronous loop, pegging CPU utilization at 100% and completely freezing the Node.js server process.
When using RBAC to apply authentication rules, the exact path (method name) matcher applies a prefix match instead of an exact match for case-insensitive matches. As a result, if a service has a method with a name that is a prefix of the name of a different method, and they have different access rules, and case-insensitive matching is used, this bug can cause improper authentication.
Netty's HttpServerCodec pairs each outbound response with an inbound request by calling pollMethod() once per response, including for 1xx informational responses. If a client pipelines an HTTP/1.1 GET carrying an Expect: 100-continue header followed by a HEAD request, the 100 Continue response consumes the queued GET method, so the subsequent 200 OK for the GET is paired with HEAD and its body is dropped, while the following 200 OK for …
Netty's HttpServerCodec pairs each outbound response with an inbound request by calling pollMethod() once per response, including for 1xx informational responses. If a client pipelines an HTTP/1.1 GET carrying an Expect: 100-continue header followed by a HEAD request, the 100 Continue response consumes the queued GET method, so the subsequent 200 OK for the GET is paired with HEAD and its body is dropped, while the following 200 OK for …
The AetherBrowser API server (scripts/aetherbrowser/api_server.py) exposes the POST /api/ops/check-email endpoint without any authentication. Any remote attacker can call this endpoint and trigger execution of the email_reader.py subprocess, which connects to configured ProtonMail or Gmail accounts via IMAP and returns email metadata (sender, subject, body snippet) in the JSON response. The server binds to 0.0.0.0:8100 by default with CORS set to allow_origins=["*"], making it reachable from any network or browser origin. …
Two locations consume an encrypted handle returned by an untrusted external party and use it without verifying that the party is ACL-authorized on it.
With $wgEmbedVideoRequireConsent disabled (not the default), the urls for videos are passed into an iframe src attribute without sanitization. When given a malformed url or id, the src attribute can be escaped via double quotes, allowing for html/javascript injection.
Overview Verified. Multiple Unrestricted Path Traversal vulnerabilities exist in the Knowns MCP docs and memory tools, allowing arbitrary file read, write, and deletion operations outside the project sandbox. The storage layer functions (Get, Create, Update, Rename, Delete) in both doc_store.go and memory_store.go concatenate user-controlled paths with filepath.Join() without any containment validation. Additionally, the docs.update action with a newPath parameter performs a file deletion via Rename(), but is classified as CapWrite …
The /home/{file_path:path} endpoint in web_client.py serves static files by directly concatenating the user-supplied file_path with the home_directory constant. There is no path traversal filtering, no path normalization check, and no authentication required. An attacker can use ../ sequences to read arbitrary files from the server filesystem.
An unauthenticated account takeover vulnerability exists in fof/oauth when the Discord OAuth provider is enabled. Discord allows an account to use an unverified email address when its phone number has been verified. During OAuth authentication, Discord may return that email address with "verified": false. Affected versions of fof/oauth did not validate this flag and passed the email address to Flarum core as trusted via provideTrustedEmail(). Flarum core could then match …
A vulnerability exists in containerd's image pull handlers where a crafted OCI image index containing deeply nested or heavily fanned-out descriptor graphs can cause unbounded CPU and memory consumption. During the PullImage operation, the recursive traversal and processing of child descriptors lack sufficient depth and breadth limits, and fail to adequately deduplicate identical descriptors. This unbounded traversal leads to excessive resource allocation. Consequently, pulling a malicious image reference can result …
A vulnerability exists in containerd's image pull handlers where a crafted OCI image index containing deeply nested or heavily fanned-out descriptor graphs can cause unbounded CPU and memory consumption. During the PullImage operation, the recursive traversal and processing of child descriptors lack sufficient depth and breadth limits, and fail to adequately deduplicate identical descriptors. This unbounded traversal leads to excessive resource allocation. Consequently, pulling a malicious image reference can result …
The vulnerability allows an attacker to bypass the HTML sanitizer by using the data-html-content attribute in the content of a SharpEditorFormField.
A Stored Cross-Site Scripting (XSS) vulnerability exists in the rich text editor due to improper sanitization of the srcdoc attribute on elements. While the underlying Symfony HtmlSanitizer correctly HTML-encodes special characters inside the attribute value (e.g., converting to <script>), the HTML specification mandates that browsers automatically decode HTML entities inside attribute values before processing them. As a result, any encoded JavaScript inside a srcdoc attribute is evaluated and executed as …
An OS command injection vulnerability exists in the PowerShell and Cmd shell wrappers provided by the CliInvoke.Specializations package (the PowershellProcessInvoker/CmdProcessInvoker invokers, and the UsePowerShell/UseCmd middleware in v3 pre-release versions). The wrappers re-run a caller-supplied target and arguments inside a shell command (pwsh -Command … / cmd /c …). In affected versions the wrapped command was delivered to the operating system as a single ProcessStartInfo.Arguments string. The OS command-line parser re-tokenizes …
An OS command injection vulnerability exists in the PowerShell and Cmd shell wrappers provided by the CliInvoke.Specializations package (the PowershellProcessInvoker/CmdProcessInvoker invokers, and the UsePowerShell/UseCmd middleware in v3 pre-release versions). The wrappers re-run a caller-supplied target and arguments inside a shell command (pwsh -Command … / cmd /c …). In affected versions the wrapped command was delivered to the operating system as a single ProcessStartInfo.Arguments string. The OS command-line parser re-tokenizes …
An argument-injection vulnerability exists in the CliInvoke package's runner factory: RunnerProcessFactory on the 2.x line and RunnerConfigurationFactory on the 3.x line. The factory joins the runner arguments, the caller's target, and the caller's arguments into a single ProcessStartInfo.Arguments string and hands it to the OS. The OS command-line parser re-tokenizes the string before the runner sees it. A double quote (") in the target or in any argument closes the …
An argument-injection vulnerability exists in the CliInvoke package's runner factory: RunnerProcessFactory on the 2.x line and RunnerConfigurationFactory on the 3.x line. The factory joins the runner arguments, the caller's target, and the caller's arguments into a single ProcessStartInfo.Arguments string and hands it to the OS. The OS command-line parser re-tokenizes the string before the runner sees it. A double quote (") in the target or in any argument closes the …
A vulnerability in ZITADEL’s Login V2 UI allowed a password-verified browser session to be reused for a new authentication request without re-checking a user’s enrolled second factor (TOTP, OTP, or U2F). An attacker who already knows valid credentials can fully authenticate to an application without completing MFA.
A vulnerability in ZITADEL Actions V1 allows an organization Action author to read files from the ZITADEL host filesystem through the JavaScript require() module loader. On common self-hosted deployments this can be chained to steal bootstrap credentials (including the Login Client PAT) and escalate from a single-tenant organization owner to instance administrator.
An uncontrolled resource consumption / algorithmic complexity vulnerability affecting any application that parses untrusted email with this library. Three independent parsing paths are super-linear in cost, so a byte-size cap on the caller side does not bound the work done. A crafted message under 2 MB can consume seconds of CPU or hundreds of megabytes to multiple gigabytes of memory (leading to an out-of-memory kill), enabling denial of service. The …
A CRLF (carriage-return / line-feed) header injection affecting any application that uses this library to build or forward MIME messages with an attacker-influenced attachment filename. Attachment filenames are interpolated into the Content-Type and Content-Disposition header values without stripping CR/LF, so a filename containing \r\n serializes as one or more additional, attacker-controlled header lines (for example a forged Bcc: that silently exfiltrates a copy of the outgoing message). The untrusted filename …
[!NOTE] This finding was identified during an agentic unsafe Rust code review performed by Gemini AI, followed by human review and verification.
xhtml-purifier does not HTML-entity-encode attribute values when serializing its sanitized output. In attributeString() (XHTMLPurifier.js, around line 148) the attribute value is concatenated directly into a double-quoted attribute without encoding. As a result, an attacker-controlled value in any allowed attribute (class, style, title, alt, src, href) can include a double-quote character to break out of the attribute and inject an additional attribute, such as a JavaScript event handler (for example onmouseover …
Reporter: Cavan Loughran, Celvex Group Inc. Summary The fix for CVE-2026-46439 (3.12.2 / 4.0.3) removed the recursive re-render loop in trestle/core/commands/author/jinja.py render_template, but the custom include tags in trestle/core/jinja/tags.py (MDSectionInclude, MDCleanInclude) still re-parse the CONTENT of an included markdown file as a Jinja2 template via Parser(self.environment, ).parse() in a plain (non-sandboxed) jinja2.Environment. Because the trestle markdown writers (ssp_io.py SSPMarkdownWriter, docs_control_writer.py DocsControlWriter) write OSCAL prose / component-description fields verbatim (the {{ …
At a glance Actor: attacker who controls the -o/–output argument to trestle author {catalog,profile,ssp}-generate (e.g. via a CI pipeline that derives the output directory from repository-controlled data) Primitive: attacker-controlled –output value reaches trestle_root / args.output write sink with only is_directory_name_allowed() (parts[0]-only task-name-collision check), not the PathSecurityValidator.validate_local_path() guard added by the CVE-2026-46345 fix Impact: arbitrary-location file write outside the trestle workspace as the process owner (8.4 High; conservative C:N variant 7.7, …
SUNEDITOR v2.47.10 appears to allow JavaScript execution through crafted namespaced HTML elements. The sanitization logic does not fully remove executable event-handler attributes from certain custom/namespaced tags. As a result, an attacker may be able to inject HTML content that executes JavaScript when the rendered element is interacted with. This behavior was observed after the changes introduced in this commit: https://github.com/JiHong88/suneditor/commit/9ed405fb0de676e56cd0e6a13c19c103ad5948d3
HTTPSession mounts a FileAdapter for the file:// scheme and inherits redirect handling unchanged from requests.Session. requests does not check for scheme downgrades or cross-protocol transitions when it follows a redirect, so any http(s) request Streamlink makes can be redirected by the remote server into file:///… and the local file is returned as the response body. GHSA-hgqw-6m45-hw5f / CVE-2026-44353 fixed direct file:// segment and playlist URIs in HLS playlists and DASH …
sprintf-js through 1.1.3 passes unbounded precision specifiers to toFixed, toExponential, and toPrecision methods without validation, causing uncaught RangeError exceptions. Attackers who control format strings can inject precision values exceeding ECMAScript limits to abort calling operations with minimal payload.
The vk-app backend accepted VK application callback data without verifying the callback signature when the auth_key parameter was omitted. Applications using this backend could treat unsigned attacker-controlled data as a verified VK identity. An attacker could choose callback fields such as viewer_id, access_token, api_id, and api_result, potentially allowing authentication as an arbitrary VK user ID. The issue affects only applications using the vk-app backend.
The Vend OAuth2 backend used only the numeric Vend user_id as the social-auth UID. When multiple Vend shops authenticate through the same application, users from different shops with the same internal Vend user ID could collide in the social-auth association table. A user from one shop could then be authenticated as the local account previously associated with the same numeric user ID from another shop. The issue affects applications using …
The LoginRadius backend did not validate OAuth state during the authentication flow. Applications using this backend were vulnerable to login CSRF. An attacker could cause a victim's browser session to complete authentication using an attacker-controlled LoginRadius token, making the victim authenticated as the attacker's LoginRadius identity. The issue affects only applications using the LoginRadius backend.
The SAML backend accepted SAML responses on the Assertion Consumer Service endpoint without verifying that they matched a previously issued AuthnRequest. Applications using SAML account association could allow an attacker with a valid account on a trusted IdP to link the attacker's SAML identity to a logged-in victim's local account. The attacker could then authenticate through SAML and gain access to the victim's account. The issue affects applications using the …
The partial-pipeline resume mechanism accepted partial_token as a bearer credential without binding it to the browser session that created it. Applications using resumable partial pipeline steps could allow an attacker to start an authentication flow, obtain a valid partial token and verification data, and cause a victim's browser to resume that attacker-controlled flow. This could authenticate the victim's browser as the attacker's account. The issue affects applications using partial pipeline …
Snipe-IT's uploaded-files API accepts XML documents and later serves them inline without applying the safe-inline allowlist used by the equivalent web controller. An authenticated user who can attach files to a supported object can upload an XSLT stylesheet and an XML document that references it through xml-stylesheet. When another authorized user opens the XML file through the API with ?inline=true, the browser applies the attacker-controlled stylesheet, which produces HTML containing …
A user with the "customfields.create" permission can store HTML/JS in a Custom Field name, which is later rendered as an asset-list column title WITHOUT escaping at app/Presenters/AssetPresenter.php line 364 ('title' => $field->name) and injected into the table header by the bundled bootstrap-table plugin. It executes for anyone who opens an asset list (e.g. /hardware), including superusers, on page load with no interaction. Since "customfields.create" can be granted to non-superusers, a …
2FA is enforced only by the web middleware group, not the api group, and the personal-access-token endpoint is in the api group. A session that has passed the password check but not the 2FA can mint a persistent API token and use it for full API access.
The rex_list component reads the SQL sort column directly from the sort GET parameter without validating it against the set of columns declared sortable via setColumnSortable(). Although the value is wrapped in backticks via escapeIdentifier() (preventing classical SQL injection), this still allows any authenticated backend user to ORDER BY any column in the query's FROM tables, including unselected sensitive columns such as password from the rex_user table, and perform error-based …
An container image that contains a environment variable with just a key and no value can trick podman into passing that variable from the host into the container. This is made worse by the fact that using an asterisk (*) will cause podman to pass all host variables into the container. So essentially a malicious image can exfiltrate all podman environment variables that are set in the session from where …
An container image that contains a environment variable with just a key and no value can trick podman into passing that variable from the host into the container. This is made worse by the fact that using an asterisk (*) will cause podman to pass all host variables into the container. So essentially a malicious image can exfiltrate all podman environment variables that are set in the session from where …
An container image that contains a environment variable with just a key and no value can trick podman into passing that variable from the host into the container. This is made worse by the fact that using an asterisk (*) will cause podman to pass all host variables into the container. So essentially a malicious image can exfiltrate all podman environment variables that are set in the session from where …
An container image that contains a environment variable with just a key and no value can trick podman into passing that variable from the host into the container. This is made worse by the fact that using an asterisk (*) will cause podman to pass all host variables into the container. So essentially a malicious image can exfiltrate all podman environment variables that are set in the session from where …
An container image that contains a environment variable with just a key and no value can trick podman into passing that variable from the host into the container. This is made worse by the fact that using an asterisk (*) will cause podman to pass all host variables into the container. So essentially a malicious image can exfiltrate all podman environment variables that are set in the session from where …
An container image that contains a environment variable with just a key and no value can trick podman into passing that variable from the host into the container. This is made worse by the fact that using an asterisk (*) will cause podman to pass all host variables into the container. So essentially a malicious image can exfiltrate all podman environment variables that are set in the session from where …
When running podman quadlet install –replace to replace a Quadlet file, if the original Quadlet is larger than the new Quadlet, the file would not be truncated and content from the original would be preserved. There is no risk of information leakage as the user already had access to the Quadlet in order to replace it, and in most cases, this would only lead to invalid Quadlet files (see https://github.com/podman-container-tools/podman/issues/29013). …
The public two-factor verification endpoint POST /check logs a user in based solely on a valid 6-digit TOTP token and a chosen user-id. It does not require — and is not bound to — a prior successful password authentication. For any account that has 2FA enabled, an unauthenticated attacker can authenticate without knowing the password, reducing the account to a single factor (a 6-digit code) that is itself brute-forceable because …
The public two-factor verification endpoint POST /check logs a user in based solely on a valid 6-digit TOTP token and a chosen user-id. It does not require — and is not bound to — a prior successful password authentication. For any account that has 2FA enabled, an unauthenticated attacker can authenticate without knowing the password, reducing the account to a single factor (a 6-digit code) that is itself brute-forceable because …
A stored cross-site scripting (XSS) vulnerability in phpMyFAQ allows any unauthenticated user (or low-privileged registered user) to inject arbitrary JavaScript that executes in an administrator's browser when they review or edit a user-submitted FAQ entry. This leads to admin account takeover via session theft. The vulnerability exists because html_entity_decode() converts HTML entities into executable HTML after strip_tags() has already passed them through, and the admin template renders the content with …
A stored cross-site scripting (XSS) vulnerability in phpMyFAQ allows any unauthenticated user (or low-privileged registered user) to inject arbitrary JavaScript that executes in an administrator's browser when they review or edit a user-submitted FAQ entry. This leads to admin account takeover via session theft. The vulnerability exists because html_entity_decode() converts HTML entities into executable HTML after strip_tags() has already passed them through, and the admin template renders the content with …
The StopWords::add() method in phpMyFAQ builds a SQL INSERT statement using sprintf() and inserts the user-supplied stop word value directly into the query string without calling the application's database escaping function on it. A sibling method, StopWords::update(), which modifies an existing stop word, correctly escapes the same kind of input. The omission is isolated to the add() (insert) code path. An authenticated administrator who can reach the stop-word management feature …
The StopWords::add() method in phpMyFAQ builds a SQL INSERT statement using sprintf() and inserts the user-supplied stop word value directly into the query string without calling the application's database escaping function on it. A sibling method, StopWords::update(), which modifies an existing stop word, correctly escapes the same kind of input. The omission is isolated to the add() (insert) code path. An authenticated administrator who can reach the stop-word management feature …
Language Servers for AWS (the aws/language-servers project) provides the Language Server Protocol implementations that power AWS developer tooling, including the Amazon Q Developer agentic chat experience, across IDEs such as VS Code, JetBrains, Visual Studio, and Eclipse. Missing symlink validation in Language Servers for AWS may allow an arbitrary file write outside of the workspace trust boundary. This may occur when a local user opens a workspace containing a maliciously …
Language Servers for AWS (the aws/language-servers project) provide the underlying language-server runtime that powers Amazon Q Developer's AI coding assistance across its IDE plugins (Visual Studio Code, JetBrains, Eclipse, and Visual Studio). Improper trust boundary enforcement in Language Servers for AWS may allow for arbitrary code execution. If a local user opens a maliciously crafted workspace, any commands within the project configuration files may be automatically executed.
langchain-nvidia-ai-endpoints versions before 1.4.2 accepted local filesystem paths as image inputs for Vision Language Model (VLM) requests. If an application passed attacker-controlled image input to ChatNVIDIA or VLM reranking APIs, an attacker could cause files readable by the application process to be read and included in the outbound request to the configured NVIDIA/NIM model endpoint.
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
Impact Type: Improper authorization leading to unauthorized movement of user funds. The x/bonds module moved funds from an address that was resolved from a DID verification method, without verifying that the resolved address belonged to the transaction signer. Affected handlers included MsgMakeOutcomePayment, MsgBuy, MsgSell, MsgSwap, and MsgWithdrawShare, as well as the batch order processor. Because any account may list an arbitrary blockchainAccountID as a verification method on a DID it …
http4s-scala-xml provides EntityDecoder[F, scala.xml.Elem] instances that parse XML message bodies. These decoders used a javax.xml.parsers.SAXParserFactory obtained from SAXParserFactory.newInstance without any security configuration. With the JDK's default settings, the parser resolves DOCTYPE declarations, external general and parameter entities, and external DTDs. An application that uses these decoders to parse untrusted XML is vulnerable to XML External Entity (XXE) attacks. An attacker can craft a request that: discloses local files readable by …
http4s-scala-xml provides EntityDecoder[F, scala.xml.Elem] instances that parse XML message bodies. These decoders used a javax.xml.parsers.SAXParserFactory obtained from SAXParserFactory.newInstance without any security configuration. With the JDK's default settings, the parser resolves DOCTYPE declarations, external general and parameter entities, and external DTDs. An application that uses these decoders to parse untrusted XML is vulnerable to XML External Entity (XXE) attacks. An attacker can craft a request that: discloses local files readable by …
http4s-scala-xml provides EntityDecoder[F, scala.xml.Elem] instances that parse XML message bodies. These decoders used a javax.xml.parsers.SAXParserFactory obtained from SAXParserFactory.newInstance without any security configuration. With the JDK's default settings, the parser resolves DOCTYPE declarations, external general and parameter entities, and external DTDs. An application that uses these decoders to parse untrusted XML is vulnerable to XML External Entity (XXE) attacks. An attacker can craft a request that: discloses local files readable by …
Users of the python-hyper/hpack library, most commonly used as downstream dependency of the python-hyper/h2 library (an HTTP/2 client and server implementation). Unbounded variable integer decoding can cause run-away computation on malformed input leading to O(n^2) runtime, effectively blocking further processing with large enough unsanitized input.
Dozzle supports per-user label filters in users.yml that are documented as an access-control boundary: "Filters are used to restrict the containers that a user can see" and "the guest user can only see containers with the label com.example.app … useful for restricting access to specific containers" (docs/guide/authentication.md). This is the mechanism operators use to give different users/tenants visibility into disjoint subsets of containers on the same host. The events stream …
Decepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK (Bring Your Own Key) deployment model, users configure their own LLM credentials to any OpenAI-compatible endpoint. Most open-source and self-deployed model providers (vLLM, SGLang, Ollama, LM Studio, text-generation-webui, etc.) do not filter special-token literals from user content in their default configurations. Those literals are …
Decepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK (Bring Your Own Key) deployment model, users configure their own LLM credentials to any OpenAI-compatible endpoint. Most open-source and self-deployed model providers (vLLM, SGLang, Ollama, LM Studio, text-generation-webui, etc.) do not filter special-token literals from user content in their default configurations. Those literals are …
Decepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK (Bring Your Own Key) deployment model, users configure their own LLM credentials to any OpenAI-compatible endpoint. Most open-source and self-deployed model providers (vLLM, SGLang, Ollama, LM Studio, text-generation-webui, etc.) do not filter special-token literals from user content in their default configurations. Those literals are …
DBHub 0.21.2 exposes an unauthenticated HTTP MCP endpoint when started with the documented HTTP transport mode, for example –transport http –port 8080. The HTTP server attempts to protect browser-origin access by checking whether the Origin hostname equals the Host hostname, then reflecting the validated Origin into Access-Control-Allow-Origin. This does not stop DNS rebinding. After an attacker-controlled hostname rebinds to a victim-accessible DBHub HTTP server, both Origin and Host can contain …
On 5 June 2026 CyberChef received a security vulnerability report from @hyuunnn detailing a vulnerability in the Series Chart operation, where malicious input could result in prototype pollution of the data structures outputted from the operation. Other operations following Series Chart could have their behaviour manipulated by the attacker-controlled prototype, for example, injecting malicious content into their HTML output. In this case, a demonstration was provided that chained the Series …
The Feed Reader front-end module passes RSS feed URLs from its configuration directly to $this->feedIo->read($url) without any scheme validation or private-IP blocklist. A backend user with module-edit permissions can configure an arbitrary URL pointing to internal network services, cloud-provider metadata endpoints, or loopback addresses, causing the server to fetch those resources unconditionally. Confirmed live: the server successfully reaches the internal MySQL database container and its own loopback Apache instance.
The Feed Reader front-end module passes RSS feed URLs from its configuration directly to $this->feedIo->read($url) without any scheme validation or private-IP blocklist. A backend user with module-edit permissions can configure an arbitrary URL pointing to internal network services, cloud-provider metadata endpoints, or loopback addresses, causing the server to fetch those resources unconditionally. Confirmed live: the server successfully reaches the internal MySQL database container and its own loopback Apache instance.
The Cline Hub dashboard server (@cline/cline-hub), launched via the cline dashboard CLI command, accepts WebSocket connections on the /browser endpoint without validating the HTTP Origin header. When ROOM_SECRET is not set—the default for local (127.0.0.1) binds—isAuthorizedBrowserRequest() returns true unconditionally, allowing any website a developer visits to open a cross-origin WebSocket to ws://127.0.0.1:8787/browser. An attacker-controlled page can then send desktopCommand frames to read workspace/session state, mutate MCP and provider settings, and—because …
In Cilium clusters using Gateway API, users with permissions to create or update namespaced HTTPRoutes can mirror HTTP traffic to any Service in any namespace, bypassing the ReferenceGrant authorization mechanism. Gateway API functionality is disabled by default.
In Cilium clusters using Gateway API, users with permissions to create or update namespaced HTTPRoutes can mirror HTTP traffic to any Service in any namespace, bypassing the ReferenceGrant authorization mechanism. Gateway API functionality is disabled by default.
Reported by @echennells (Eric Chennells). Migrated from public issue #191 to a private advisory. Affected: @bsv/wallet-toolbox / -client / -mobile. Verified in 2.1.21 and 2.1.21-parity-fix.2; the relevant code is the same at current HEAD Summary: When createAction runs against a remote StorageClient, the storage server returns the outputs to build. buildSignableTransaction takes each non-change output's lockingScript from the storage response and signs it, without comparing it to the lockingScript the …
Reported by @echennells (Eric Chennells). Migrated from public issue #191 to a private advisory. Affected: @bsv/wallet-toolbox / -client / -mobile. Verified in 2.1.21 and 2.1.21-parity-fix.2; the relevant code is the same at current HEAD Summary: When createAction runs against a remote StorageClient, the storage server returns the outputs to build. buildSignableTransaction takes each non-change output's lockingScript from the storage response and signs it, without comparing it to the lockingScript the …
Reported by @echennells (Eric Chennells). Migrated from public issue #191 to a private advisory. Affected: @bsv/wallet-toolbox / -client / -mobile. Verified in 2.1.21 and 2.1.21-parity-fix.2; the relevant code is the same at current HEAD Summary: When createAction runs against a remote StorageClient, the storage server returns the outputs to build. buildSignableTransaction takes each non-change output's lockingScript from the storage response and signs it, without comparing it to the lockingScript the …
The default Rsdoctor report HTTP server started by @rsdoctor/rspack-plugin binds to all network interfaces (0.0.0.0) and serves a POST /api/data/key endpoint with no authentication and wildcard CORS (Access-Control-Allow-Origin: *). Any network-adjacent or remote attacker can send a single unauthenticated request to retrieve the full source code of all compiled JavaScript modules (moduleCodeMap), serialized build configuration (configs), error details, and other sensitive build metadata. This server is enabled by default in …
Setting readonly = true on the execute_sql tool does not make the connection read-only. The connectors are written to set PostgreSQL default_transaction_read_only=on (and open SQLite in readOnly mode), but that code is gated on a config value that is never populated, so it never runs. The only thing left enforcing read-only is a classifier that inspects the first keyword of each statement. Any SELECT that writes or has side effects …
Denial of service via memory exhaustion. Affects all callers who streamed compressed responses relying on the chunk size — explicit (iter_bytes(chunk_size=…)) or the default — to bound memory. The decoder ignored that bound, so a chunk could be far larger than requested and a single compressed response could overflow memory. import gzip, zapros
Who is impacted: Any application using Zapros to make HTTP requests to untrusted servers Applications that follow redirects to attacker-controlled hosts Attack vector: A malicious HTTP server returns a response with many chained content encodings. When the client attempts to decode, it creates a deeply nested decompression chain consuming excessive resources.
WPGraphQL 2.19.0 contains an authorization bypass in the updatePost mutation. An authenticated WordPress Contributor can change one of their own draft posts to PUBLISH despite lacking the publish_posts capability. The same mutation also permits the Contributor to modify their own previously published posts despite lacking edit_published_posts and failing WordPress's object-level edit_post capability check. This bypasses the standard WordPress editorial workflow. The WordPress REST API correctly rejects the equivalent operations for …
Wire's Swift runtime (Wire SPM/CocoaPods product, implemented by wire-runtime-swift) did not reject a negative LENGTH_DELIMITED field length while skipping an unknown protobuf group. A crafted 10-byte protobuf payload could cause ProtoReader.skipGroup() to read a length-delimited field whose varint decodes to a negative Int32. That negative value was then passed to ReadBuffer.readData(count:). ReadBuffer checked only that the requested read did not go past the end of the buffer. It did not …
An unauthenticated JCR-SQL2 injection exists in the Smart Content category filter of the 2.x content query builder. Category IDs supplied through the public ?categories= query parameter are only trimmed and are then concatenated directly into a JCR-SQL2 WHERE clause, without the numeric validation that the equivalent tag filter already performs. Any public page that renders a Smart Content element with category filtering enabled evaluates this parameter, so no authentication or …
A limited server-side Twig template injection issue was identified in Solspace Freeform for Craft CMS. In affected versions, submitted form field values could be evaluated as Twig in certain rendering contexts. Unauthenticated users could submit Twig expressions through public Freeform forms and have those expressions evaluated when the submitted values were rendered back into the form response. Confirmed impact includes disclosure of limited server/environment information, such as the PHP version …
A stored cross-site scripting (XSS) vulnerability exists in REDAXO CMS 5.x. When an administrator attempts to delete a media file that is referenced by a Media Manager effect, the warning message rendered in the backend includes the type's name field without HTML escaping. An attacker with access to the Media Manager addon can store an XSS payload as a type name; the payload executes in the browser of any administrator …
The mediapool sync page (sync.php) renders filenames from the /media filesystem directory directly into HTML without applying rex_escape() (i.e., htmlspecialchars). Any file placed in the media directory whose filename contains HTML metacharacters will execute JavaScript in the browser of any backend user who views the sync page.
The rex_api_install_package_update API function (install addon) does not override requiresCsrfProtection(), which defaults to false in the base class rex_api_function. Any authenticated admin can therefore be tricked via a CSRF attack into silently triggering a package update from the REDAXO package server.
ReactPress API list endpoints build TypeORM QueryBuilder conditions using unsanitized HTTP query parameter names as SQL column identifiers (e.g. article.${key}). TypeORM parameterizes values but not column names, allowing unauthenticated attackers to inject SQL through crafted query string keys.
The Classic portlet (plone.app.portlets.portlets.classic) used its user-supplied template/macro fields to build a TALES path expression that was then evaluated by the TAL path() helper. Because the value was interpreted as a full TALES expression, a user able to add or edit a Classic portlet could supply a crafted value that escapes simple path traversal and is evaluated as arbitrary code. This is exploitable by any authenticated user who can configure …
When creating content with a very large title or description field, Plone can become unresponsive. Even if the content loads, it can be hard to edit or delete, because the UI is unwieldy.
When creating an Image or File with a very large filename, Plone can become unresponsive. Even if the content loads, it can be hard to edit or delete, because the UI is unwieldy.
A user who can save a telemetry screen (permission system_set) can embed JavaScript in a screen BUTTON widget. The BUTTON widget eval()s the stored button text in the browser when the button is activated, and screens are shared content rendered to other users in the scope. As a result, an attacker's stored JavaScript executes in a different operator's authenticated session — a stored, cross-user XSS (not self-XSS). The payload runs …
COSMOS reads configuration from a user-writable overlay (targets_modified/) before the read-only plugin-installed targets/ tree, and the config subsystem executes code on those files: ConfigParser renders every file as ERB by default, a GENERIC_READ_CONVERSION / GENERIC_WRITE_CONVERSION block is evaluated as code by GenericConversion (Ruby and Python), and the Script Runner suite analysis requires a procedure file. An authenticated user can write into targets_modified/ below the admin tier (the storage-upload endpoint exempts …
An authenticated user can execute arbitrary operating system commands on the openc3-cosmos-cmd-tlm-api service. The pypi_url setting is interpolated, unescaped, into a command line that is run through a shell backtick when a plugin is installed. Shell metacharacters in the setting value are executed by /bin/sh.
moquette is reachable by untrusted MQTT clients (anonymous by default), so every byte from any client, including pre-authentication, is untrusted. This is a memory-safe JVM: the ceiling is authorization/ACL bypass + denial of service + cross-session integrity, not RCE (I did not find one and do not claim one). Audited at commit da7f719a6bab9829d520b5838e13ea7b1f9be3ef, module broker/.
The /csp endpoint accepts unauthenticated JSON reports and logs user-controlled values directly to stdout using print() without escaping control characters. A remote attacker can include ANSI/VT100 escape sequences in fields such as blocked-uri or document-uri. When the logs are viewed in an ANSI-capable terminal, these sequences can manipulate the displayed output (e.g., clear the screen, hide text, or inject misleading messages), affecting the integrity of operator-facing logs.
Location: core/kapp/market/market.go — Buy() (approx. L281–436) Severity: High The native marketplace enforces an IsClaimed guard in Claim (market.go:752), CancelOrder (market.go:1125), and orderEscrowAmount (market.go:251), but not in Buy. Marketplace escrow is virtual: the market KApp account never custodies currency. A bid burns funds from the bidder (bidderAcc.SubFromBalance, ~L378), and refunds/payouts mint to the recipient (AddToBalance, e.g. the prior-bidder refund at ~L349). Soundness depends on each order's CurrentBid being paid out exactly …
Location: core/kapp/validators/validators.go:201 (Register), (genesis/checking/nodesSetupChecker.go:73). core/consensus/slot/bls/subslotStartSlot.go:165 core/consensus/…/headerSignatureVerify.go:123 (Create(…)).
The unauthenticated WebSocket endpoint GET /subscribe is registered open: true by default (config/node/api.yaml) and lets a remote, unauthenticated client exhaust the node's memory and goroutines. Because the REST API runs IN-PROCESS with the node — network/api/api.go Start(…) ends with ws.Run(kleverFacade.RestAPIInterface()) — exhausting/killing the API process takes down the entire node, including its P2P and consensus participation. No API key, account, stake, or funds are required. Three compounding, independently-exploitable gaps stack …
Component: Elasticsearch indexer (indexer/) Primary location: indexer/common.go:2395-2407 (serializedDataForUpdateAccounts) Entry point: SetAccountName native transaction (contract type 12) — core/process/transaction/txProcess.go:688
An unauthenticated client can connect to GET /log, send an arbitrary logger profile as the first WebSocket message, and mutate the node's global logging configuration before receiving live logs from the process. I confirmed this against a local validator built from this repository: an unauthenticated client set the global log level to *:NONE, the node accepted the profile, and the node stopped emitting normal slot logs while the WebSocket connection …
The VM built-in function KleverUpdateAccountPermission (registered always-active, creator.go:381-390 / core/vmconstants.go:234) rewrites an account's entire permission set. Its authorization check uses vmInput.RecipientAddr attacker-controlled instead of the authenticated vmInput.CallerAddr. The sibling handler kleverChangeOwnerAddress.go:86 uses vmInput.CallerAddr correctly, so the safe pattern exists in-repo; this handler deviates. The native transaction path (txProcess.go:833) is safe it uses tx.GetSender(). Mechanism: Wrong variable: CallerAddr is never referenced in the handler; auth is contractHasValidPermission(target.GetPermissions(), RecipientAddr), which returns true …
The JLine3 HISTORY_IGNORE variable is converted into a Java regular expression with only partial escaping. As a result, regex metacharacters other than * and : are passed through to the regex engine. A crafted value such as (a+)+b can cause catastrophic backtracking each time a command line is added to history, hanging the reader thread at high CPU.
When regex search mode is enabled in the JLine3 nano editor, the user-supplied search term is compiled directly as a Java regular expression with no timeout or backtracking bound. A crafted pattern such as (a+)+b can hang the editor session thread at high CPU, causing a denial of service for that session.
The JLine3 built-in grep command wraps the user-supplied regular expression with .* before compiling it with Java's backtracking regex engine. This amplifies catastrophic backtracking and allows a short pattern such as (a+)+b to hang the command thread on non-matching input. In environments that expose the JLine shell to remote users, this is a denial-of-service issue.
The Jawn parser before 1.6.1 is vulnerable to a denial of service attack via untrusted input.
The Jawn parser before 1.6.1 is vulnerable to a denial of service attack via untrusted input.
The Jawn parser before 1.6.1 is vulnerable to a denial of service attack via untrusted input.
AsyncParser can be forced to perform O(n^2) work on the length of the input. When a single JSON token arrives across many small chunks, each absorb call rescans the incomplete token from the start.
AsyncParser can be forced to perform O(n^2) work on the length of the input. When a single JSON token arrives across many small chunks, each absorb call rescans the incomplete token from the start.
AsyncParser can be forced to perform O(n^2) work on the length of the input. When a single JSON token arrives across many small chunks, each absorb call rescans the incomplete token from the start.
Incomplete cleanup vulnerability in Apache Tomcat allows a malformed request to potentially (depends on timing) cause one request from another user to fail. This issue affects Apache Tomcat: from 11.0.19 through 11.0.25, from 10.1.53 through 10.1.59, from 9.0.116 through 9.0.121. Users are recommended to upgrade to version 11.0.26, 10.1.60 or 9.0.122, which fix the issue.
Incomplete cleanup vulnerability in Apache Tomcat allows a malformed request to potentially (depends on timing) cause one request from another user to fail. This issue affects Apache Tomcat: from 11.0.19 through 11.0.25, from 10.1.53 through 10.1.59, from 9.0.116 through 9.0.121. Users are recommended to upgrade to version 11.0.26, 10.1.60 or 9.0.122, which fix the issue.
Improper handling of length parameter inconsistency vulnerability in Apache Tomcat allows WebSocket message smuggling when per-message-deflate is used. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.25, from 10.1.0-M1 through 10.1.59, from 9.0.0.M1 through 9.0.121. The following versions were EOS at the time the CVE was created but are known to be affected: from 8.5.0 through 8.5.100, from 7.0.56 through 7.0.109. Other unsupported versions may also be affected. Users are …
Improper handling of length parameter inconsistency vulnerability in Apache Tomcat allows WebSocket message smuggling when per-message-deflate is used. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.25, from 10.1.0-M1 through 10.1.59, from 9.0.0.M1 through 9.0.121. The following versions were EOS at the time the CVE was created but are known to be affected: from 8.5.0 through 8.5.100, from 7.0.56 through 7.0.109. Other unsupported versions may also be affected. Users are …
The anonymous front-end action formie/submissions/submit (SubmissionsController::actionSubmit) trusted a client-supplied submissionId when loading an incomplete submission, with no session binding, ownership check, or edit token validation. An unauthenticated attacker could enumerate sequential submission IDs and overwrite or hijack another user's in-progress (multi-page / save-for-later) submission. Tampered data could be persisted and forwarded via notifications and integrations when the submission was completed. Completed submissions were not affected (isIncomplete=true filter). This is an …
The control panel action formie/sent-notifications/get-resend-modal-content (SentNotificationsController::actionGetResendModalContent) performed only requireAcceptsJson() and loaded a SentNotification by request id without permission or object-level authorization checks. Any authenticated user who could invoke the action could enumerate notification IDs and read full email content — including recipient headers and the complete HTML body containing submitted form data (PII) — without formie-accessSentNotifications or equivalent permission. Sibling actions in the same controller enforced authorization.
The control panel action formie/integrations/form-settings (IntegrationsController::actionFormSettings) was reachable by any authenticated user without the appropriate form integration permissions. The action applied request-supplied settings to a fully configured integration via setAttributes($settings, false), allowing an attacker to overwrite outbound host properties (e.g. apiUrl) while the server sent stored API keys or OAuth tokens to the attacker-controlled host. The remote response was returned in the JSON body (non-blind SSRF). This is an incomplete …
Elysia v1.4.28 is vulnerable to denial-of-service attacks due to CPU exhaustion in the form data normalization code. Elysia uses getAll to retrieve value from FormData. It is called directly relative to the total number of key-value pairs in the form data. The total amount of work the for loop has to do grows quadratically, so doubling the number of unique key-value pairs quadruples the amount of work. In the above …
Authentication Bypass by Alternate Name vulnerability in Apache Tomcat allowed the security constraints for any WebSocket endpoint to be bypassed. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.25, from 10.1.0-M1 through 10.1.59, from 9.0.0.M1 through 9.0.121. The following versions were EOS at the time the CVE was created but are known to be affected: from 8.5.0 through 8.5.100, from 7.0.43 through 7.0.109. Other unsupported versions may also be affected. …
Authentication Bypass by Alternate Name vulnerability in Apache Tomcat allowed the security constraints for any WebSocket endpoint to be bypassed. This issue affects Apache Tomcat: from 11.0.0-M1 through 11.0.25, from 10.1.0-M1 through 10.1.59, from 9.0.0.M1 through 9.0.121. The following versions were EOS at the time the CVE was created but are known to be affected: from 8.5.0 through 8.5.100, from 7.0.43 through 7.0.109. Other unsupported versions may also be affected. …
The PATCH /api/settings endpoint writes the entire request body to persistent settings without a field whitelist. An authenticated user can set security-critical fields that are not meant to be modifiable here — notably requireLogin. Setting requireLogin: false disables authentication for the whole application, exposing all protected routes (e.g. /api/keys, /api/providers) to unauthenticated access.
The Kiro API-key validation endpoint builds an upstream URL using a user-controlled region value. By supplying a crafted region such as kiro-canary.local:8443#, an authenticated attacker can cause 9router to send the Kiro validation request to an attacker-controlled host under the constructed codewhisperer.<region> hostname. The request forwards the submitted Kiro API key as an Authorization: Bearer header.
9router validates image URLs by resolving the host before fetching, but the later server-side fetch performs a separate DNS resolution. An attacker-controlled DNS name can resolve to a public IP during validation and then rebind to an internal Docker/private IP during the fetch. This allows the server-side image prefetch to reach internal-only HTTP services (SSRF).
9router treats local loopback requests as trusted and allows access to /v1/* without an API key. In a documented/common reverse-proxy deployment where nginx forwards public traffic to the backend via 127.0.0.1, external non-Origin requests are misclassified as local. This allows unauthenticated access to /v1 APIs such as /v1/models, and may allow abuse of configured upstream provider credentials depending on the enabled providers.
wlc could send an unscoped API token to an unintended server when run inside a directory tree containing attacker-controlled project configuration. If .weblate, .weblate.ini, or weblate.ini defines an API url, and the user supplies a token with WLC_KEY or –key without also pinning the URL, wlc would send the token to the project-configured URL. Impacted users are those running wlc in untrusted repositories, pull request checkouts, or directories with untrusted …
Multiple authorization vulnerabilities in Unleash admin API, including a critical missing await that completely bypasses a permission check.
The change-request approval HTML email template renders fields as raw HTML. User who can create a change request can inject HTML into approval notification emails. I was not able to confirm Enterprise platform is using vulnerable code nor that it doesn't sanitize input.
The clone-feature endpoint supports copying features across projects, but it does not verify that the caller can access the source project. A user with create permissions in one project can clone a feature from another project they cannot read and then inspect the copied configuration. This vulnerability cannot be confirmed without Enterprise access. Report is based on a circumstantial evidence in the open-source repository.
Unleash scopes write permissions per project and per environment: a user with the UPDATE_FEATURE_STRATEGY permission on project A is supposed to be able to mutate activation strategies only within project A. The endpoint POST /api/admin/projects/:projectId/features/:featureName/environments/:environment/strategies/set-sort-order violates this. The RBAC middleware authorizes the request against the :projectId taken from the URL, but the handler then writes the strategy IDs supplied in the request body directly to the database by primary key, …
Traefik's BasicAuth middleware coalesces concurrent credential checks through a singleflight.Group to avoid hashing the same password many times at once. Since v3.6.11 the deduplication key was built from the submitted password plus the stored secret, so it depended on server state: a non-existent username collapsed onto one shared key while each configured username produced its own. Under attacker-controlled concurrency, a probe request arriving inside a leader request's in-flight window is …
Traefik's BasicAuth middleware coalesces concurrent credential checks through a singleflight.Group to avoid hashing the same password many times at once. Since v3.6.11 the deduplication key was built from the submitted password plus the stored secret, so it depended on server state: a non-existent username collapsed onto one shared key while each configured username produced its own. Under attacker-controlled concurrency, a probe request arriving inside a leader request's in-flight window is …
It's possible to enumerate users through a timing oracle. In other words: I can easily check if a username exists or not by observing the timing differences between logins.
Tinyauth's login rate-limit bookkeeping can enter a global lockdown mode when its in-memory login-attempt map reaches 256 distinct identifiers. Because unauthenticated POST /api/user/login requests for unknown usernames are recorded in this same map, a remote unauthenticated attacker can submit 257 unique bogus usernames and cause valid credentials for unrelated users to be treated as locked until auth.loginTimeout expires. This was confirmed against the stable v5.0.7 release. With default configuration, auth.loginTimeout …
tinyauth is a forward-auth service: a reverse proxy (Traefik/Caddy/nginx/Envoy) calls GET /api/auth/<proxy> on every request and only forwards the request upstream if tinyauth returns 200. tinyauth decides which per-app access rules apply by looking up the forwarded hostname (the app) in its ACL set — the static apps: config and/or Docker labels. Each app can restrict access with users.allow / users.block, oauth.whitelist, oauth.groups / ldap.groups, and ip.allow. These allowlists are …
validateUser() in backend/src/authentication/providers/mysql/auth-provider-mysql.service.ts returns immediately when the supplied login/email does not match any account, without ever calling comparePassword(): async validateUser(loginOrEmail: string, password: string, ip?: string, scope?: AUTH_SCOPE): Promise { let user: UserModel try { user = await this.usersManager.findUser(loginOrEmail, false) } catch (e) { … } if (!user) { this.logger.warn(…) return null // <– comparePassword() is never reached here } return await this.usersManager.logUser(user, password, ip, scope) } comparePassword() (backend/src/common/functions.ts) already contains …
In the vulnerable version, the sync diff endpoint accepted a user-controlled regex pattern through pathFilters and compiled it into a RegExp without complexity validation. The resulting regex was then executed synchronously against relative file paths during diff generation. A catastrophic-backtracking pattern, such as ^(a+)+b, can block the affected Node.js event loop when evaluated against a worst-case path shape. In a single-process deployment, this can make the server unavailable to other …
POST /api/auth/token authenticates with username and password only, then calls getTokens(), which returns unrestricted Bearer access and refresh JWTs without checking whether the account has TOTP 2FA enabled. An attacker who already knows valid credentials for a 2FA-enabled account can bypass 2FA in a single request. The parallel login endpoint (POST /api/auth/login) correctly enforces 2FA by calling setCookies(user, res, true), which gates on user.twoFaEnabled when server-side TOTP is enabled.
The Store API v3 endpoint PATCH /api/v3/store/carts/:id/associate binds a guest cart to the authenticated caller without verifying possession of that cart. It locates the cart by prefixed ID only — current_store.carts.where(user: [nil, current_user]).find_by_prefix_id!(params[:id]) — and omits the authorize!(:update, @cart, cart_token) check that every other action in the controller performs via CartResolvable. Because prefixed IDs are a reversible Sqids encoding of the auto-increment primary key (obfuscation, not a token), an authenticated …
The WebSocket transport allocates a buffer from the frame payload length before validating its size, which can lead to an unauthenticated DoS.
The stream parser allocates the SIP body buffer from the Content-Length header before validating its size, which can lead to an unauthenticated DoS.
RequestFilteringHttpAgent / RequestFilteringHttpsAgent block requests to private IPs, but the blocking happens via a synchronous throw inside createConnection() for literal private-IP hostnames (e.g. 169.254.169.254, 127.0.0.1). Node.js's http.request / http.get expects createConnection to emit an error asynchronously; a synchronous throw instead escapes the caller's req.on('error', …) handler entirely and becomes an unhandled exception that crashes the process.
PSDImage.composite() (and .numpy()) allocate the output image buffer from the PSD's header geometry (width × height × channels × depth, and per-layer rectangles) before validating those values against the actual file contents. A tiny crafted PSD declaring huge dimensions causes a multi-gigabyte allocation. Critically, composite() then returns a (black) image with only a warning, no exception is raised, so a caller cannot detect or guard against it.
OpenCVE contains a Server-Side Request Forgery (SSRF) vulnerability in the notification testing functionality for both Webhook and Slack integrations. An authenticated user with permission to configure notification channels can trigger test requests to arbitrary HTTP(S) endpoints. Insufficient validation of target destinations allows requests to be sent to hosts reachable from the OpenCVE server, including internal network resources, localhost interfaces, and cloud metadata services. Successful exploitation could allow an attacker to …
When running in the highly privileged recovery mode, OpenBao was vulnerable to a timing attack against the single recovery token. This allowed an attacker to extract the recovery token and use it to perform operations against the OpenBao instance, including reading or modification of data.
When a policy operator has written capabilities = ["deny"] on a path with a trailing wildcard but allowed a broader list operation (e.g., a deny on secrets/metadata/restricted/* but allowed list on secrets/metadata/*), OpenBao would incorrectly allow the operation. This did not impact other operation types.
During certain error conditions, OpenBao Agent's exec rendering mode will incorrectly write secrets specified in env_template to stdout. This primarily happens when num_retries is met. This vulnerability is original to Vault and was reported via the OpenBao security mailing list.
A vulnerability in Nuclei's workflow template loader allows file: protocol templates to execute without the -file flag, bypassing a security gate that is meant to prevent local file reads on the scanner host. Affected Component The issue is in the workflow template loading path. The main template loader enforces the -file gate for file-protocol templates, but the workflow loader did not apply the same check when resolving templates referenced by …
A vulnerability in Nuclei's JavaScript MySQL client library allows arbitrary local file reads that bypass the -allow-local-file-access (-lfa) sandbox restriction. Affected Component The issue is in the nuclei/mysql JavaScript library used by javascript: protocol templates. The MySQL client connection logic did not enforce the local file access sandbox when handling the allowAllFiles DSN parameter used by LOAD DATA LOCAL INFILE. Description Without -lfa, Nuclei confines local file access to the …
A vulnerability in Nuclei's DAST/fuzz expression evaluation path allows a malicious target server to trigger disclosure of scanner-host environment variables when the -env-vars / -ev option is explicitly enabled. This is an incomplete fix for CVE-2026-41645 / GHSA-jm34-66cf-qpvr. The original fix hardened expressions.Evaluate() to be single-pass within one call, but did not address callers that invoked evaluation multiple times on substituted output in the DAST/fuzz pipeline. Affected Component The issue …
A vulnerability in Nuclei's DAST template loading path allows unsigned code: protocol templates to execute, bypassing the cryptographic signature requirement that is meant to prevent arbitrary command execution from untrusted templates. Affected Component The issue is in the template loader's DAST loading branch. When -dast is enabled and a template contains a fuzzing: block, the loader accepted the template through a code path that omitted the unsigned-code-template signature check present …
A vulnerability in the Goja JavaScript engine used by Nuclei's javascript: protocol allows arbitrary native code execution on the scanner host when running untrusted JavaScript templates. Affected Component The issue is in the Goja JavaScript runtime embedded in Nuclei's JavaScript protocol (pkg/js/). An out-of-bounds heap write in the engine can be exploited to achieve native code execution during template evaluation. Description Nuclei uses the Goja engine to execute javascript: protocol …
The @novu/js In-App Inbox renderer passes a notification's redirect.url to window.open() with no URL-scheme validation. The value originates from a notification's call-to-action and is delivered to the recipient verbatim. An authenticated organization member (or any holder of the environment API key) creates a v1 in-app workflow whose step CTA stores cta.data = { url: "javascript:<payload>", target: "_self" }. The v1 message-template cta.data field is a Mongoose Mixed type, so the …
What kind of vulnerability is it? Who is impacted? It has two related instances that share the same root cause: a user-controlled model field is assigned verbatim to a form field's help_text, which is rendered with Django's |safe filter (render_field.html), bypassing auto-escaping. In both cases the script executes in the browser of any user who opens an affected create or edit form, including administrators and superusers. Because the payload runs …
This is an authorization bypass that escalates into unauthorized server-side job execution. Primary impact - self-approval: The approver checks (approver-group membership, change permission on the object under review, one-response-per-user) are enforced only in the approve/deny actions on ApprovalWorkflowStage. The generic ApprovalWorkflowStageResponse create endpoint enforces none of them, so a user holding only extras.add_approvalworkflowstageresponse can POST approved responses directly and drive a stage past its min_approvers threshold with no legitimate approver …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
MPXJ used the default configuration when creating a DocumentBuilder instance, which leaves doctype declarations enabled, when parsing the XML content of the ZTIMEINTERVALS column from a Merlin project SQLite file. This would allow a carefully crafted XML payload to read an arbitrary file. However, although an arbitrary file can be read, the way the resulting parsed XML is processed by MPXJ means that the data it contains is unlikely to …
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
When reading a suitably crafted PRX or STX file, MPXJ can be made to write files to arbitrary locations in the file system.
A malicious client can force the server to pay ~5x the normal transaction fee. This dramatically increases operational costs and completely destroys the profit margin on low-cost items. Please provide a way to share the PoC in .zip format to you. Thank you!
A malicious client can force the server to pay ~9.4x the normal transaction fee. This dramatically increases operational costs and completely destroys the profit margin on low-cost items.
When the SDK spawns a sandbox, the msb sandbox child process receives the full network configuration as an inline –network-config <json> command-line argument, and any per-sandbox environment as repeated –env KEY=VALUE arguments. On Linux a process's arguments are world-readable via /proc/<pid>/cmdline, and on both Linux and macOS they are visible to other local processes via ps. Because the network configuration carries the real secret values used for host-side secret substitution, …
The UserTokenMiddleware extracts URLs from X-Atlassian-Jira-Url and X-Atlassian-Confluence-Url HTTP headers and passes them directly to API client constructors without any SSRF validation.
In the documented multi-user HTTP deployment (–transport streamable-http with global operator Atlassian credentials), sooperset/mcp-atlassian exposes all tools to unauthenticated network clients, and the upload_attachment tool reads an attacker-supplied file_path with no path validation. Chained, an unauthenticated network attacker reads arbitrary files on the MCP server host (e.g. /proc/self/environ → the operator's Atlassian API token + .env secrets, ~/.ssh/id_rsa, /etc/passwd) by uploading them to an attacker-chosen page/issue and reading them back.
The SSRF protection for header-based authentication uses a validate-then-use pattern vulnerable to DNS rebinding. validate_url_for_ssrf resolves the hostname via DNS and checks that the resolved IP is globally routable. However, the actual HTTP request happens later, during which the DNS record may have changed to point to an internal IP (127.0.0.1, 169.254.169.254, etc.). The SSRF redirect hook only validates redirect targets, not the initial connection.
_make_ssrf_safe_hook() blocks HTTP redirects to private/internal IPs by validating the Location header before the client follows a 3xx response. The problem is that this hook is only attached in one of three authentication branches — the header-PAT path. Basic auth and OAuth branches skip it entirely, so if the connected Atlassian server returns a redirect to something like http://169.254.169.254/, the requests session follows it without complaint. This is an incomplete …
The SSRF protection in validate_url_for_ssrf() can be bypassed with a URL containing a backslash before userinfo-like syntax. Affected code: parsed = urlparse(url) hostname = parsed.hostname … ip_error = _check_ip_address(hostname) … dns_error = _check_dns_resolution(hostname) Payload: http://127.0.0.1:6666@www.baidu.com For this input, urllib.parse.urlparse() treats the hostname as: www.baidu.com Therefore, validate_url_for_ssrf() validates www.baidu.com instead of 127.0.0.1. However, the downstream request made through the Atlassian client / requests.Session reaches the local service: http://127.0.0.1:6666/%5C@www.baidu.com/rest/api/2/myself This allows an …
The OAuth 2.0 setup wizard's local callback HTTP server reflects the error query parameter directly into an HTML response without any sanitization or encoding. An attacker can craft a malicious callback URL containing JavaScript in the error parameter that executes in the victim's browser when the setup wizard is running. The server binds to all network interfaces (0.0.0.0), making it accessible from the local network rather than just localhost.
The upload_attachment functions in both the Jira and Confluence modules accept a user-controlled file_path parameter and open the specified file for reading without calling validate_safe_path(). An authenticated MCP client can supply an arbitrary path such as /etc/passwd or /proc/self/environ, causing the server process to read and transmit the file's contents to the remote Atlassian instance as an attachment. This is an incomplete fix relative to GHSA-xjgw-4wvw-rgm4: the download_attachment and download_issue_attachments …
The confluence_upload_attachment and confluence_upload_attachments MCP tools accept a file_path parameter and do not validate that the path is confined to an allowed directory before opening the file. An attacker who can call these tools can read any file accessible to the MCP server process (SSH keys, .env files, API credentials) and exfiltrate it by uploading it to Confluence. Note: The Jira upload_attachment mixin method in jira/attachments.py has the same missing …
mcp-atlassian exposes an MCP tool confluence_upload_attachment whose file_path argument is passed directly to open(file_path, "rb") without any path validation. An attacker able to invoke the tool can read arbitrary files readable by the server process and exfiltrate them into a multipart upload directed at an attacker-controlled Confluence host. In the default streamable-http transport the server binds 0.0.0.0 with no built-in authentication, making this remotely exploitable without credentials. This is the …
When OAuth tokens are saved, MCP Atlassian always writes a plaintext fallback copy under ~/.mcp-atlassian/oauth-<client_id>.json. The fallback file is created with the process default umask rather than restrictive permissions. In this environment the file was created as mode 0664, exposing access and refresh tokens to same-group local users and any process that can read the home directory.
The mcp-atlassian server exposes an MCP tool (confluence_upload_attachment and the Jira attachment variant) that accepts an arbitrary server-side file path and opens it for upload without any path validation. When the server is deployed in HTTP transport mode (streamable-http or sse), a remote, unauthenticated attacker can supply attacker-controlled Atlassian service headers (X-Atlassian-Confluence-Url / X-Atlassian-Confluence-Personal-Token) to redirect the upload to an attacker-controlled endpoint, then pass an arbitrary file_path (e.g. /etc/passwd, ~/.env, …
mcp-atlassian is a popular community MCP server wrapper exposing Jira / Confluence to MCP clients. Operators commonly restrict the surface to a small allowlist of projects/spaces via the JIRA_PROJECTS_FILTER and CONFLUENCE_SPACES_FILTER environment variables, which the README documents as the principal mechanism for limiting attacker-controlled MCP clients (= prompt-injected LLM agents) to the operator's intended subset of the workspace. We identified three distinct sub-bugs that let attacker-controlled queries read content from …
MCP Atlassian exposes Jira and Confluence attachment upload tools that accept arbitrary local filesystem paths and upload those file contents to Atlassian. In HTTP or multi-user deployments, an MCP caller who can invoke write tools can cause the server to read any file accessible to the MCP process and send it to Jira/Confluence as an attachment.
The OAuth token fallback file storage in OAuthConfig._save_tokens_to_file() creates token files containing access tokens, refresh tokens, and cloud IDs with default filesystem permissions (typically 0644 on Linux, world-readable). Any local user on a shared system can read these files to obtain full Atlassian API credentials, enabling unauthorized access to the victim's Jira and Confluence data.
The path traversal fix introduced in v0.17.0 (GHSA-xjgw-4wvw-rgm4) is incomplete. validate_safe_path() is called without an explicit base_dir, defaulting to os.getcwd(). In standard container deployments the process CWD is the application directory (e.g. /app), so paths within that directory, including the application's own Python source modules, pass validation without raising an exception. An attacker can overwrite a module file and achieve remote code execution on the next process restart. Versions >= …
mcp-atlassian supports HTTP/SSE deployment for persistent, remote, and multi-user use. In that mode, write-capable Jira and Confluence attachment flows accept caller-controlled server-local file paths and read those paths without an upload-source allowlist or base directory restriction. An authenticated caller, or a prompt-injected agent with tool-call capability, can cause the MCP server process to read a local file it can access and attach that file into a Jira issue or Confluence …
ENABLED_TOOLS and TOOLSETS filters are enforced at tools/list time only. tools/call dispatches from the full unfiltered tool registry (73 tools). Any user with access to the server endpoint that knows a tool name can invoke it directly. Tool names are not secret since mcp-atlassian is open source. Any direct JSON-RPC call bypasses the restriction entirely. READ_ONLY_MODE is not affected: it has dual enforcement at list time (_list_tools_mcp) and call time …
The Jira and Confluence attachment upload tools accept caller-controlled file path parameters and read those paths from the MCP server's local filesystem before uploading the file as an Atlassian attachment. In local stdio deployments, this can expose files readable by the user's MCP process. In documented HTTP/SSE or streamable-http deployments, the impact is higher: any MCP client that is allowed to invoke write/upload tools can cause the server process to …