Infinite loop denial-of-service in bytebot 0.0.1's Agent Execution Workflow can be triggered remotely by a low-privileged attacker, exhausting server resources and disrupting service availability. The product is end-of-life and no longer maintained by bytebot-ai, meaning no official patch will be released. A public proof-of-concept exploit is available on GitHub, lowering the bar for exploitation despite the moderate CVSS 4.0 score of 5.3.
Remote denial-of-service in Open5GS HSS up to version 2.8.0 allows a low-privilege Diameter Cx peer to crash the Home Subscriber Server process by sending a UAR, MAR, SAR, or LIR request containing an empty User-Name or Public-Identity AVP, triggering a reachable assertion (CWE-617) in src/hss/hss-cx-path.c. A proof-of-concept is publicly available via GitHub issue #4675, and the upstream fix exists as commit c9abe09421eb99bbf1cd7862a3d375e58a4eb9e4; no formally tagged patched release version has been independently confirmed. No CISA KEV listing is present, constraining the assessed threat primarily to actors with internal Diameter network access.
Remote code injection in provectus kafka-ui up to version 0.7.2 allows a low-privileged authenticated attacker to execute arbitrary Groovy code on the server via the smart filter test endpoint. The vulnerability resides in the executeSmartFilterTest function of MessagesController.java, where user-supplied Groovy expressions are evaluated without adequate sandboxing or restriction. A publicly available proof-of-concept exploit exists, no vendor patch has been released, and the project maintainers have not responded to the responsible disclosure report.
SQL injection in itsourcecode Library Management System 1.0 allows remote authenticated attackers to manipulate database queries via the ID parameter in editbooks.php. The vulnerability is classified as CWE-89 (Improper Neutralization of Special Elements used in an SQL Command) and affects the book editing function, potentially exposing or corrupting the underlying database. A public proof-of-concept exploit exists, raising the operational risk for any unpatched deployment.
SQL injection in itsourcecode Sales and Inventory System 1.0 exposes the customer edit endpoint to database manipulation by authenticated network attackers who can control the ID parameter in /pages/cust_edit.php. The CVSS 4.0 base score of 2.1 reflects limited impact across confidentiality, integrity, and availability, though a public proof-of-concept exploit exists on GitHub. No active exploitation by CISA KEV has been confirmed; the low score and limited deployment footprint constrain real-world priority despite the public exploit.
Path traversal in Faveo Helpdesk versions 2.0.0 through 2.0.3 enables remote authenticated administrators to delete arbitrary files on the server by manipulating the `data1` argument passed to PHP's `unlink()` function within the Logo Handler component (`SettingsController.php`). Exploitation is limited to accounts holding high privileges, constraining realistic impact despite network reachability. Publicly available exploit code exists via a GitHub issue report; the vendor has not responded to the disclosure as of time of analysis.
OS command injection in sworddut mcp-ffmpeg-helper affects the handleToolCall function in src/tools/handlers.ts across all three released versions (0.1.0, 0.1.1, 0.2.1), allowing a local low-privilege attacker to execute arbitrary OS commands by manipulating the format argument. Publicly available exploit code exists via a GitHub issue, and the vendor has not responded to the coordinated disclosure, leaving all known versions unpatched. This is not listed in CISA KEV; exploitation requires local access, which substantially limits real-world exposure.
Reflected cross-site scripting in Piwigo 16.3.0 public authentication pages allows a remote attacker to inject arbitrary JavaScript into a victim's browser by planting a malicious `lang` cookie value that is reflected unencoded through the application's fatal error handler. The vulnerability spans four public-facing PHP files (identification.php, password.php, register.php, profile.php) and is exploitable without authentication. A publicly available proof-of-concept exists; vendor-released patch 16.4.0 is available and exploitation is not confirmed in CISA KEV, though the AC:H complexity requirement meaningfully constrains real-world risk.
OS command injection in TanStack devtools-vite 0.7.0 allows a local low-privileged attacker to execute arbitrary operating system commands by manipulating the `packageName` argument to the `installPackage` function within the Development Devtools Event Bus component. Exploitation is constrained to local access with high attack complexity, yielding limited confidentiality, integrity, and availability impact on the vulnerable system only. A public proof-of-concept was disclosed via GitHub issue #464 prior to any vendor response; no patch has been released as of this analysis.
Null byte injection in Netty's MqttEncoder affects applications using versions prior to 4.1.137.Final and 4.2.17.Final that forward user-controlled input into MQTT UTF-8 string fields - specifically client identifiers, will topics, usernames, and PUBLISH topic names. Because the encoder performs no null byte validation, an attacker with adjacent-network access and low-level privileges can embed the prohibited U+0000 character, causing downstream MQTT brokers to misparse field boundaries and potentially bypass topic-based ACLs or misidentify client sessions. No public exploit has been identified at time of analysis, and the vulnerability has not been added to the CISA KEV catalog.
OAuth token refresh in hex.pm (Elixir's package registry) fails to re-validate organizational membership when issuing new access tokens, allowing users removed from a private organization to retain read access to its packages. Affected are hex.pm deployments from 2025-10-10 through 2026-08-24. An authenticated former member can repeatedly refresh a token carrying an explicit repository:<org> or docs:<org> scope - bypassing revocation entirely for up to 30 days - because CDN edges authorize access from token claims without querying the database. No public exploit exists and this is not in CISA KEV.
Heap-based buffer overflow in Open5GS 2.8.0's HSS S6a Authentication-Information-Request handler allows network-adjacent authenticated Diameter peers to corrupt process heap memory by supplying a malformed Visited-PLMN-Id AVP with an incorrect octet-string length. The vulnerable function `hss_ogs_diam_s6a_air_cb` in `src/hss/hss-s6a-path.c` passes raw AVP data to cryptographic key derivation (`ogs_auc_kasme`) without first validating that the PLMN identifier length matches the expected 3-byte constant `OGS_PLMN_ID_LEN`. No active exploitation is confirmed (no CISA KEV listing), but a proof-of-concept signal is present in the CVSS 4.0 threat metric (E:P); an upstream fix commit is available though a formally tagged patched release is not independently confirmed.
Authorization bypass in Dolibarr ERP up to versions 18.0.10, 22.0.5, and 23.0.3 allows any low-privileged authenticated user to read or overwrite another user's private notes by manipulating the numeric `id` parameter in `/user/note.php`. The root cause (CWE-639 - Authorization Bypass Through User-Controlled Key) is a missing identity check before the permission evaluation: all requests were evaluated under the caller's self-write right rather than distinguishing whether the target user matched the caller. A proof-of-concept exists (CVSS 4.0 E:P), though no active exploitation has been confirmed via CISA KEV. Patches are available in versions 23.0.4 and 24.0.0.
Out-of-bounds read in Open5GS 2.8.0 PCRF Rx AA-Request handler allows remote low-privileged attackers with Diameter Rx interface access to send crafted AA-Request messages containing null or incorrectly-sized Framed-IP-Address or Framed-IPv6-Prefix AVPs, triggering memory reads beyond allocated boundaries. The patch commit c18dc6938bf63cc7374315d3dca303d92066e746 adds missing null and length validation for both AVP types; a tagged release incorporating it has not been independently confirmed. No active exploitation is recorded in CISA KEV, though the CVSS 4.0 E:P flag indicates proof-of-concept code is known to exist.
Improper authorization in Open5GS 2.8.0's AMF UEContextReleaseRequest path handler allows remote low-privileged attackers to trigger unauthorized UE context release operations in the 5G core network. The flaw, tagged as an authentication bypass (CWE-285), affects the Access and Mobility Management Function - a security-critical component governing device connectivity and session management in 5G deployments. A proof-of-concept exploit exists (E:P in CVSS 4.0 supplemental), and no CISA KEV listing was identified at time of analysis.