CVE Feed

In the Linux kernel, the following vulnerability has been resolved: HID: uclogic: fix use-after-free of inrange_timer on remove uclogic_remove() cancels the pen in-range timer and then stops the device: timer_delete_sync(&drvdata->inrange_timer); hid_hw_stop(hdev); timer_delete_sync() only guarantees the timer is idle at that instant. uclogic_raw_event_pen() keeps delivering pen reports until hid_hw_stop() stops the transport several lines later, and every report with pen->inrange == UCLOGIC_PARAMS_PEN_INRANGE_NONE re-arms the timer: mod_timer(&drvdata->inrange_timer, jiffies + msecs_to_jiffies(100)); A report landing between the timer_delete_sync() call and the transport teardown in hid_hw_stop() re-arms inrange_timer after it was cancelled. uclogic_remove() then returns and the devm drvdata is freed, while hid_hw_stop() has already freed the input device drvdata->pen_input points at, so when the timer fires ~100 ms later uclogic_inrange_timeout() dereferences freed memory -- a use-after-free in timer-softirq context. Swapping the two calls is not a fix: stopping the device first frees drvdata->pen_input via hidinput_disconnect() while the timer may still be pending, so a timer already armed before removal fires on the freed input device in the window before timer_delete_sync() runs. Use timer_shutdown_sync() before hid_hw_stop() instead. It cancels the timer, waits for a running callback while pen_input is still valid, and prevents any further re-arming -- a later mod_timer() from an in-flight report is silently ignored -- so the timer is provably dead before hid_hw_stop() frees the inputs. This is the ordering the timer core documents for this "timer re-armed from another path" teardown case.

In the Linux kernel, the following vulnerability has been resolved: HID: hyperv: validate initial device info bounds The Hyper-V synthetic HID host supplies SYNTH_HID_INITIAL_DEVICE_INFO messages that contain a HID descriptor followed by the report descriptor bytes. mousevsc_on_receive_device_info() trusts bLength and wDescriptorLength without checking that the received packet contains both byte ranges. A malformed host or backend message can therefore make the guest read past the received VMBus packet while copying the report descriptor. Pass the received initial-device-info size into the parser and reject descriptor lengths that exceed the packet. Impact: A malicious Hyper-V host or backend can crash a guest by sending a short initial device-info message with an oversized HID report descriptor length.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: hci_event: fix LE list UAF on reset hci_cc_reset() clears the LE accept and resolving lists without taking hdev->lock. Other command-complete handlers serialize updates to these lists with that lock, and the debugfs readers hold it while walking them. This permits the reset completion and a debugfs read to interleave as follows: hci_rx_work debugfs reader ----------- -------------- lock hdev->lock fetch current entry list_del(entry) kfree(entry) read entry fields The reader then dereferences a freed list entry and may follow its stale next pointer. KASAN reported: BUG: KASAN: slab-use-after-free in white_list_show+0x15f/0x180 Read of size 1 at addr ffff8881015dab16 by task poc/95 Call Trace: white_list_show+0x15f/0x180 seq_read_iter+0x3ff/0x1190 seq_read+0x267/0x3d0 vfs_read+0x177/0xa20 ksys_read+0xf7/0x1c0 Allocated by task 91: hci_bdaddr_list_add+0x1a6/0x3a0 hci_cc_le_add_to_accept_list+0xab/0x140 hci_cmd_complete_evt+0x26c/0x9a0 hci_event_packet+0x454/0xb20 hci_rx_work+0x293/0x730 Freed by task 90: kfree+0x131/0x3c0 hci_bdaddr_list_clear+0xd8/0x160 hci_cc_reset+0x28a/0x370 hci_cmd_complete_evt+0x26c/0x9a0 hci_event_packet+0x454/0xb20 hci_rx_work+0x293/0x730 Take hdev->lock around both list clears. This matches the existing mutation and traversal locking convention.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: hci_event: validate LE Set CIG Parameters response The Command Complete dispatch validates only the fixed part of the LE Set CIG Parameters response. After that part is pulled from the skb, hci_cc_le_set_cig_params() trusts num_handles and reads each entry in the trailing handle array. Matching num_handles against the command's num_cis does not guarantee that the response contains the advertised handles. A truncated response from a malfunctioning controller can therefore make the handler read beyond the skb data. Validate that the remaining skb data contains all advertised handles. Include this in the existing response validation so malformed responses also follow the established CIG failure handling.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: hci_sync: Fix accept list UAF during suspend hci_update_event_filter_sync() walks hdev->accept_list while sending a synchronous HCI command for each remote-wakeup device. The suspend path holds hdev->req_lock, but accept-list updates are serialized by hdev->lock. Consequently, remove_device() can free the current list entry during the controller wait. The following interleaving causes the use-after-free: hci_update_event_filter_sync() remove_device() fetch accept-list entry hci_set_event_filter_sync() wait for controller response hci_dev_lock() list_del() kfree() hci_dev_unlock() read the freed list.next KASAN reported: BUG: KASAN: slab-use-after-free in hci_suspend_sync+0x835/0x910 Read of size 8 at addr ffff88810bec8440 by task kworker/0:1/10 Workqueue: events vhci_suspend_work Call Trace: hci_suspend_sync+0x835/0x910 hci_suspend_dev+0x182/0x450 process_one_work+0x661/0x1090 worker_thread+0x45b/0xd10 Allocated by task 86: hci_bdaddr_list_add_with_flags+0x1a8/0x400 add_device+0x381/0x820 hci_sock_sendmsg+0x1033/0x1ea0 Freed by task 91: kfree+0x131/0x3c0 remove_device+0x429/0xb70 hci_sock_sendmsg+0x1033/0x1ea0 Snapshot the remote-wakeup addresses under hdev->lock. Release the lock before sending HCI commands. Clear the controller event filter before building the snapshot, and skip allocation and the second list traversal when there are no matching entries. This preserves the original filter and scan-state updates without retaining an accept-list node across a controller wait.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: ISO: zero the sockaddr before returning it in getname iso_sock_getname() fills a struct sockaddr_iso in place and returns its size without clearing it first, so bytes it does not write are copied to user space from the kernel stack. The getsockname(2) and getpeername(2) paths both run through do_getsockname(), which hands getname() an uninitialized sockaddr_storage on the stack and copies back up to the number of bytes getname() returns, so the driver has to initialize every byte it accounts for. Two ranges are left uninitialized: - struct sockaddr_iso is 10 bytes but only 9 are written (family, iso_bdaddr, iso_bdaddr_type), leaking the trailing pad byte on every call. - for a broadcast peer (BIS_LINK or PA_LINK) the returned length grows by sizeof(struct sockaddr_iso_bc), but only bc_sid, bc_num_bis and bc_bis are filled; bc_bdaddr and bc_bdaddr_type, the first 7 bytes of that structure, are never written. An unprivileged process can open a BTPROTO_ISO socket and reach the pad leak with getsockname(); the broadcast leak needs an established BIS/PA connection. l2cap and rfcomm already memset their sockaddr in getname for the same reason; do the same here.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: MGMT: reject HCI_CMD_SYNC params_len above 255 mgmt_hci_cmd_sync() checks that the message length agrees with params_len but puts no upper bound on it. params_len is __le16 while the parameter length in the HCI command header is a u8: struct hci_command_hdr { __le16 opcode; __u8 plen; } __packed; hci_cmd_sync_alloc() assigns one to the other: hdr->plen = plen; if (plen) skb_put_data(skb, param, plen); so a params_len of 256 leaves plen at 0 while all 256 bytes are still appended. The frame handed to the driver then declares no parameters and carries 256 of them. On a length framed transport such as H:4 the controller takes the trailing bytes as the start of the next packet. The mgmt socket MTU is HCI_MAX_FRAME_SIZE, so params_len can reach about 1KB this way. Commit 03f1700b9b4d ("Bluetooth: MGMT: reject malformed HCI_CMD_SYNC commands") only made params_len agree with the message length, a value that fits the message but not the header field is still accepted. Reject params_len that does not fit the header field.

In the Linux kernel, the following vulnerability has been resolved: Bluetooth: hci_aml: validate firmware segment lengths aml_download_firmware() reads two lengths from the firmware header and uses them to build pointers before checking that the header and segment data are present. A truncated or inconsistent firmware image can make the driver read past firmware->data while constructing TCI commands. Reject images shorter than the header and ensure that the ICCM and DCCM ranges fit within the loaded firmware before downloading either segment.

In the Linux kernel, the following vulnerability has been resolved: futex: Avoid private hash use-after-free on final put futex_private_hash_put() drops the reference to fph before evaluating fph->mm for wake_up_var(). futex_ref_put() enables preemption again before returning. If that put drops the final reference and the task is preempted, another task can pivot to the replacement hash and free the old hash after an RCU grace period. The first task then reads fph->mm from the freed allocation when it resumes. KASAN reports a slab-use-after-free in futex_private_hash_put(), with the read at offset 24 in a freed kmalloc-512 allocation. The allocation and free stacks point to futex_hash_allocate() and the RCU free path, respectively. Load the mm pointer while the fph reference is still held and pass the saved value to wake_up_var(). wake_up_var() uses the pointer as a waitqueue key and does not dereference the mm through it.

IBM App Connect Enterprise 13.0.1.0 through 13.0.8.1, and 12.0.1.0 through 12.0.12.28 and IBM Integration Bus for z/OS 10.1.0.0 through 10.1.0.7 could allow a remote attacker to cause a denial of service due to an infinite loop.

IBM UCD - IBM UrbanCode Deploy 7.2 through 7.2.3.25, and 7.3 through 7.3.2.20 and IBM UCD - IBM DevOps Deploy 8.0 through 8.0.1.15, 8.1 through 8.1.2.8, and 8.2 through 8.2.2.1 IBM DevOps Deploy / IBM UrbanCode Deploy (UCD) is susceptible to an formation disclosure vulnerability when processing redacted property values. If a deployment is configured with a secure property that starts with certain non-ASCII characters, the redaction engine may fail to mask subsequent ASCII secure values embedded inside unsecure properties. An authenticated user with permissions to view deployment request details could exploit this flaw via the UI or API to view sensitive values in plain text that should otherwise be redacted.

IBM App Connect Enterprise 13.0.1.0 through 13.0.8.1, and 12.0.1.0 through 12.0.12.28 and IBM Integration Bus for z/OS 10.1.0.0 through 10.1.0.7 SAP Adapter is vulnerable to an XML external entity (XXE) attack.

Acunetix 25.11.251107123 for Windows contains a local privilege escalation vulnerability in the Web Vulnerability Scanning Engine (wvsc.exe) that allows low-privileged local attackers to execute arbitrary code as SYSTEM by exploiting a missing hardcoded directory path for OpenSSL-related files. Attackers can create the missing directory, place a malicious file at the expected path, and cause the SYSTEM-level wvsc.exe process to load and execute it, resulting in full privilege escalation.

CP4BA - IBM Enterprise Records could allow a local attacker to obtain sensitive information due to the use of a broken or risky cryptographic algorithm.

Onyx 4.6.6 fails to properly restrict access to custom tool credentials stored in custom_headers, allowing any authenticated user to read admin-defined API keys. Attackers with basic authentication can call GET /tool/{tool_id} or GET /tool endpoints to retrieve plaintext authorization headers and third-party API credentials, then use them to directly access upstream APIs.

jina-ai reader contains a server-side request forgery vulnerability where URL validation is performed only on the initial request but not re-applied to subsequent redirect hops. Attackers can craft a public URL that redirects to internal network addresses or cloud metadata endpoints, allowing the server to fetch and return the target's response body to the attacker.

Turso through 0.8.0-pre.8 contains an out-of-bounds read vulnerability in the table-leaf page reader that uses an attacker-controlled cell-count field without bounds validation. Attackers can craft a malicious database file with a modified cell count value to trigger an index-out-of-bounds panic when querying, causing denial of service in any application that opens untrusted database files.

Documenso 2.17.0 contains an access control vulnerability in the PDF-serving endpoint that fails to validate document visibility settings. Attackers with low privileges can read restricted documents within their team or cross-tenant by leveraging missing ownership validation on document data identifiers.

SadTalker contains an OS command injection vulnerability in the video muxing process where uploaded audio filenames are interpolated into ffmpeg commands without proper escaping. Attackers can upload audio files with shell metacharacters in the filename to break out of quoted arguments and execute arbitrary system commands when video generation occurs.

FastChat contains an authentication bypass vulnerability in the /register_worker endpoint that allows unauthenticated attackers to register arbitrary worker addresses and perform server-side request forgery. Attackers can register malicious workers under victim model names to intercept user prompts, images, and responses, or probe internal network ports across the worker mesh.

LaVague 0.2.35 contains a remote code execution vulnerability in PythonFromMarkdownExtractor.extract_as_object that evaluates untrusted language model output derived from web page content. Attackers can inject malicious Python code through web pages using indirect prompt injection to execute arbitrary code on the operator's host without review.

Chatbot UI contains an authorization bypass vulnerability in the retrieval endpoint that allows authenticated attackers to access private file content belonging to other users by supplying arbitrary file UUIDs. The endpoint uses a service-role Supabase client that bypasses row-level security and fails to validate file ownership, enabling attackers to retrieve indexed content chunks from victim files through crafted POST requests.

Nightingale (n9e), as of commit 8362cbe (main branch, confirmed 2026-08-27), contains a server-side request forgery vulnerability in the isPublicIP function in aiagent/tools/http.go, the SSRF guard for the http_fetch AI-agent tool. The function only unwraps standard IPv4-mapped (::ffff:a.b.c.d) IPv6 addresses before checking them against the forbidden-range list, and does not classify 6to4 (2002::/16), NAT64 (64:ff9b::/96, 64:ff9b:1::/48), or deprecated site-local (fec0::/10) addresses. On a dual-stack or NAT64-enabled host, an attacker able to supply a URL to the http_fetch tool can bypass the guard by encoding a forbidden IPv4 address (such as the cloud instance-metadata endpoint 169.254.169.254) in one of these IPv6 forms to reach internal or metadata services.

MegaParse 0.0.55 contains an unauthenticated server-side request forgery vulnerability in the POST /v1/url endpoint that fetches caller-supplied URLs server-side. Attackers can supply internal service URLs or metadata endpoints without authentication to read their responses directly from the JSON response.

Plandex 2.2.1 contains a path traversal vulnerability in the ApplyFiles function that allows attackers to write files outside the project directory. Attackers can influence model output through poisoned repository files or attacker-controlled context to write to arbitrary locations like shell rc or cron files, achieving code execution.

llmware 0.4.6 contains an SQL injection vulnerability in the collection-database layer (llmware/resources.py) where filter and lookup values are directly string-interpolated into SQL WHERE clauses without parameterization or escaping, in both the SQLite and PostgreSQL backends. The filter validator only checks keys against an allow-list and never sanitizes values. Attacker-controlled filter values reaching the public API via Library.block_lookup and Query.text_query_with_custom_filter / text_query_by_author_or_speaker can neutralize the intended filter to disclose rows the caller was scoped out of (cross-document/cross-collection disclosure); on PostgreSQL the flaw permits boolean- and UNION-based SQL injection.

TEN Framework 0.11.71 contains unauthenticated arbitrary file read and write vulnerabilities in the TMAN Designer file-content API endpoints. Attackers can submit POST and PUT requests to the /api/designer/v1/file-content endpoints to read arbitrary files or write malicious content to system paths, enabling code execution through authorized_keys, cron files, or executable graph files.

surya 0.22.1 screenshot server contains an unauthenticated arbitrary file read vulnerability in the /info, /page, and /process routes that accept raw file_path parameters. Attackers can read any image or PDF file on the host by supplying arbitrary file paths to Image.open or pypdfium2.PdfDocument, obtaining rendered contents as base64 and using /info as an existence oracle.

ms-swift 4.5.2 contains a server-side request forgery vulnerability in the swift deploy OpenAI-compatible API that fetches multimodal media URLs without validation or redirect filtering. Unauthenticated attackers can supply arbitrary image_url, audio_url, or video_url parameters to make the server issue requests to internal services and cloud metadata endpoints.

AgentScope through 2.0.7.post1 contains a path traversal vulnerability in LocalWorkspace.add_skill that copies arbitrary server directories into the agent workspace via an unconfined source path parameter. Attackers can supply any directory path in the skill_path request parameter to copy files into the skills directory, making them accessible through the workspace skill listing.

marker through 2.0.0 contains a path traversal vulnerability in the FastAPI /marker/upload handler that fails to sanitize the file.filename parameter. Unauthenticated attackers can supply filenames containing directory traversal sequences to write arbitrary files to any location or delete existing files on the system.

Dub contains an open redirect vulnerability in the redir_url query parameter that is accepted on every short link without validation or domain allowlist enforcement. Attackers can append the redir_url parameter to any short link to redirect visitors to arbitrary external URLs through the trusted Dub domain, bypassing destination blacklists and potentially enabling phishing attacks with link cloaking enabled.

OWL's DocumentProcessingToolkit contains a server-side request forgery vulnerability in the extract_document_content tool that fetches caller-supplied URLs with no scheme, host, or IP filtering. Attackers can inject malicious URLs through prompt injection to make the server fetch internal resources, with responses returned to the agent context.

aider (aider-chat) automatically loads a .aider.conf.yml configuration file from the root of the git repository it is launched in. A crafted repository can set test-cmd (executed at startup) or lint-cmd (executed on the first file edit), which aider runs through a shell (subprocess with shell=True) without any user confirmation, LLM interaction, or API key. Consequently, a user who clones and runs aider inside an attacker-supplied repository achieves arbitrary command execution on their machine. The behavior is long-standing and was confirmed on 0.86.3.dev (current main).

LLaMA-Factory contains a server-side request forgery vulnerability in the OpenAI-compatible API multimodal media URL handler that allows unauthenticated attackers to bypass SSRF validation. The check_ssrf_url guard validates URLs once but requests.get follows redirects and re-resolves DNS without re-validation, enabling attackers to use HTTP redirects or DNS rebinding to access internal addresses and cloud metadata endpoints.

zerox 1.1.20 contains an OS command injection vulnerability in the file download mechanism where the temporary file extension derived from document URLs is interpolated unsanitized into shell commands executed by poppler utilities. Attackers can craft document URLs with malicious file extensions containing command substitution syntax to execute arbitrary OS commands before document processing occurs.

QAnything 2.0.0 contains an authentication bypass vulnerability in the /api/local_doc_qa/get_file_base64 and /api/local_doc_qa/get_doc endpoints that allows unauthenticated attackers to access any uploaded file or document. Attackers can enumerate file identifiers through unauthenticated endpoints and retrieve base64-encoded files or parsed document chunks without ownership verification to disclose cross-tenant knowledge base content.

tokenizers (Hugging Face) is affected by an out-of-bounds buffer access in BpeBuilder::build (tokenizers/src/models/bpe/model.rs). When loading a tokenizer.json via Tokenizer::from_file/from_str, the builder sizes a scratch buffer to the longest vocabulary key, then writes each concatenated merge rule into it. A merge whose concatenated token exceeds the longest vocabulary key overruns the buffer, which Rust turns into a panic that aborts the process in Rust and FFI embeddings. This occurs at load time with no encoding required, so an attacker who supplies a crafted tokenizer.json can cause a denial of service. A secondary defect at the same location can cause a usize underflow (panic in debug, potential memory corruption in release) when continuing_subword_prefix is set and a merge token is shorter than the prefix. Observed in version 0.23.1.

potpie through 2.0.0 fails to verify user ownership on the POST /conversations/{conversation_id}/code-changes/sync endpoint. Authenticated attackers can write arbitrary file changes into other users' conversations by supplying their conversation IDs, allowing unauthorized modification of pending changes.

Xinference (affected commit 4a94832, v3.x) contains an unauthenticated arbitrary-path file read vulnerability in the POST /v1/models/llm/auto-register endpoint, which accepts a caller-supplied model_path parameter without authentication or path confinement. The endpoint reads and parses config.json, tokenizer_config.json, and chat_template.jinja files at the supplied path and reflects the parsed content back to the caller, allowing an unauthenticated attacker to probe the server filesystem and extract content of files with those names in any directory.

xiaobei through 5.5.2 fails to implement authentication or signature validation on webhook endpoints, allowing unauthenticated attackers to inject arbitrary messages into the agent pipeline. Attackers can publish malicious messages via the /webhook_worktool handler and exploit unvalidated media URL fetching to perform server-side request forgery against internal services.

OGX (formerly Llama Stack, affected at commit fbe8e0f) contains an unauthenticated server-side request forgery vulnerability in the OpenAI-compatible POST /v1/responses endpoint. MCP tool definitions accept a server_url parameter (along with headers and authorization values) that is fetched server-side without destination validation; the existing validate_url_not_private() guard used for other URL inputs is not applied to server_url. On the default starter configuration, which runs without authentication, a remote unauthenticated attacker can cause the server to open connections to arbitrary internal addresses (including cloud metadata endpoints such as http://169.254.169.254/) and forward attacker-supplied headers and bearer tokens to those destinations.

Bruno versions through 4.1.0 fail to validate file paths in request body declarations, allowing attackers to read arbitrary local files by using parent-directory traversal segments. When a collection is executed, attackers can craft a request with a body:file path containing ../ sequences that resolve outside the collection directory, causing the application to read and exfiltrate arbitrary files to attacker-controlled endpoints.

Chroma 1.5.9 fails to validate maximum bounds on HNSW index parameters max_neighbors, ef_construction, and ef_search in collection-create requests. Unauthenticated attackers can supply arbitrarily large parameter values to exhaust server memory and cause denial of service during index compaction.

Aim 3.29.1 remote tracking server fails to authenticate requests and dispatches arbitrary methods through getattr without allowlist validation. Unauthenticated attackers can register clients, instantiate Repo resources, and invoke arbitrary methods to read experiments or delete runs.

Marqo 2.26.0 contains a server-side request forgery vulnerability in the add_documents endpoint that allows unauthenticated attackers to trigger requests to arbitrary URLs by supplying malicious media field values. Attackers can exploit download_image_from_url and fetch_content_sample functions which lack destination filtering and host validation to access internal services and cloud metadata endpoints.

excel-mcp-server 0.1.8 fails to enforce path confinement in stdio mode when EXCEL_FILES_PATH is unset, allowing attackers to read and write arbitrary files. Attackers can supply unchecked file paths to read and write tools to access any file accessible to the process.

cli-mcp-server 0.2.5 contains a command allowlist bypass vulnerability in the _validate_command_with_operators function when ALLOW_SHELL_OPERATORS is enabled. Attackers can use shell command substitution syntax like $(...) or backticks to execute non-allowlisted commands that bypass the ALLOWED_COMMANDS validation check.

Trigger.dev versions before 4.5.2 fail to validate environment membership during run replay operations, allowing authenticated attackers to inject task runs into arbitrary environments. Attackers can replay their own runs into other organizations' or projects' environments to consume victim resources and pollute run history.

Trigger.dev before 4.5.2 contains a server-side request forgery vulnerability in webhook alert channel delivery URLs that are fetched without validation or SSRF protection. Authenticated users with organization membership can create alert channels with URLs targeting internal services and metadata endpoints, allowing the server to issue POST requests to restricted resources.