Path traversal in changedetection.io 0.60.7 and earlier allows remote attackers to read limited files outside the intended screenshot directory by manipulating the filename argument in the Screenshot Handler's static_content route in changedetectionio/flask_app.py, with no authentication or user interaction required per the vector (PR:N/UI:N). Exploitation is meaningfully constrained by high attack complexity (AC:H) and a 'difficult' exploitability rating, and the impact is confidentiality-only and low (VC:L), so this is a moderate-priority issue rather than a critical one. Publicly available exploit code exists, but there is no CISA KEV confirmation of active exploitation, and the vendor was contacted early about the disclosure but did not respond.
A timing side-channel in the password comparison routine (check_password in changedetectionio/flask_app.py) of dgtlmoon changedetection.io through version 0.60.7 allows remote, unauthenticated attackers to infer the shared application password character-by-character by measuring response-time differences. Exploitation is rated difficult (CVSS AC:H) and publicly available exploit code exists, though the practical signal is heavily degraded by network jitter, TLS/reverse-proxy overhead, and interpreter noise; only confidentiality of the single shared login secret is at risk, with no integrity or availability impact. No vendor patch has been identified at time of analysis, and the vendor did not respond to the disclosure.
Session fixation in Mstfakts College-Management-System (Front-end/server.php) allows a remote unauthenticated attacker to plant a session identifier that the application fails to regenerate at login, so the attacker inherits the victim's authenticated session once the victim signs in. Exploitation hinges on user interaction - the victim must complete an authenticated login through the attacker-influenced session - which is the main limiting factor in an otherwise trivially reachable flaw, and the assessed impact is limited to low confidentiality and integrity loss with no availability effect (CVSS 3.1 AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:L/A:N; vendor/NVD base 2.1). Publicly available exploit code exists (referenced from the project's GitHub issue #8), but there is no evidence of active exploitation; this is a low-priority, high-noise finding.
Reflected cross-site scripting (CWE-79) in sfturing's hosp_order open-source hospital ordering project, affecting commits up to 627f426331da8086ce8fff2017d65b1ddef384f8, allows remote unauthenticated attackers to execute arbitrary script in a victim's browser by crafting a malicious 'Search' argument to Public Search Handlers in HospitalController.java. Exploitation requires the victim to open a crafted request or URL (user interaction), and publicly available exploit code exists via the project's GitHub issue #118. The project has not responded to the report, so no vendor patch is available; real-world risk is tempered by the project's near-zero deployment footprint.
Unleash deployments prior to 8.0.3 render the change-request approval email template with Mustache triple-stash syntax, so the attacker-controlled changeRequestTitle, requesterName, and requesterEmail values are inserted into notification emails as raw, unescaped HTML. An authenticated project member who can create change requests (CVSS 4.0 vector AV:N/AC:L/PR:L; assessed CVSS 3.1 AV:N/AC:L/PR:L/UI:R) can therefore place forged links, tracking content, or visually altered text into an approver's inbox, although the approver must open the email and mail clients block script execution, so the realistic outcome is HTML content spoofing rather than code execution. No evidence of active exploitation (not in CISA KEV) and no public exploit code identified at time of analysis, though the vendor advisory documents manual reproduction steps; severity is genuinely low (CVSS 2.1) and the vendor-released patch is Unleash 8.0.3.
Arbitrary file upload in College-Notes-Gallery (rolling release up to commit 8c1cf3d98f30982d069c88ca172612c001eb39f6) lets authenticated remote attackers manipulate the 'image' parameter at /dashboard/userprofile.php?section=admin1 to bypass intended file-type restrictions and store arbitrary files on the server. Publicly available exploit code exists, but there is no KEV confirmation in the provided data, and the vendor did not issue a patch. The independent assessment rates this as a low real-world priority because valid credentials are required, the project is obscure with a tiny install base, and no official fix exists; the resulting impact can nevertheless be high if uploaded files are web-accessible and executable.
Authenticated SQL injection in Yonyou U8cloud 5.x lets a low-privilege OpenAPI user inject arbitrary SQL through the 'operator' argument of the /u8cloud/openapi/so.saleorder.sendaudit endpoint, potentially exposing or altering data held in the application's backend database. Exploitation is remote and requires no user interaction, but does require a valid low-privilege OpenAPI account and network reachability to the OpenAPI component (CVSS 4.0 PR:L, confirmed by the CVSS:4.0 vector), so purely unauthenticated attack is not established by the available data. Publicly available exploit code exists - a Go proof-of-concept targeting this exact endpoint has been published on GitHub - and the vendor was contacted early but did not respond, so no vendor-released patch has been identified at time of analysis; the CVSS 4.0 base score of 5.3 is modest largely because all three impact dimensions were rated Low, whereas an independent assessment rates confidentiality impact as High (CVSS:3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:N).
OS command injection in Moonshot AI's Kimi Code (versions 0.31.0 and earlier) lets an attacker-controlled .mcp.json file in an untrusted workspace be auto-spawned by the MCP Configuration Loader before any trust confirmation, executing arbitrary OS commands on the developer's machine. Exploitation requires the victim to open or operate on the malicious workspace (user interaction) and to run a pre-0.31.1 version; publicly available exploit code exists, but there is no confirmed active exploitation (no CISA KEV). Although the vendor rates this CVSS 4.0 at only 2.1, an independent assessment scores it 7.8 (CVSS 3.1, AV:L/AC:L/PR:N/UI:R, C:H/I:H/A:H) because CWE-78 command injection inside an AI agent's configuration loader normally yields full code execution.
Path traversal in changedetection.io 0.60.7 and earlier lets an authenticated low-privilege user read files outside the intended content directory by manipulating the filename argument handled by the static_content function in changedetectionio/flask_app.py (visual_selector_data component). Publicly available exploit code exists and the disclosure is documented by VulDB, but the maintainer did not respond, so no vendor-released patch has been confirmed; exploitation requires a valid low-privilege session (PR:L) and no user interaction. Impact is confidentiality-only - arbitrary file read with no file write, code execution, or availability effect - which keeps the CVSS 5.3 rating moderate and the value of the flaw dependent on which files the traversal can reach.
Stored cross-site scripting in changedetection.io up to and including 0.55.8 lets a malicious or compromised monitored web page inject markup that is rendered in the Visual Selector UI, because the selector string is written into the DOM with innerHTML rather than as text. Exploitation requires an authenticated user who has added a page they do not control, plus that user opening the Visual Selector for the affected watch, so impact is confined to the victim's own browser session and is limited to integrity of the displayed page. Publicly available exploit code exists, but the issue is not confirmed actively exploited (CISA KEV) and independent assessment rates it low priority given the authentication and user-interaction prerequisites; upgrading to 0.60.1 resolves it.
Cloudreve instances running versions prior to 4.18.0 fail to enforce write-scope authorization on two admin tool routes, so an attacker who already holds a valid OAuth application or API key scoped to Admin.Read can exceed that token's read-only boundary. Concretely, the misconfigured wopi GET and mail POST endpoints let such an authenticated token holder (PR:H in the CVSS vector) probe configured WOPI service endpoints and send arbitrary test email through the server's SMTP configuration, with no user interaction required. Impact is bounded to information probing and SMTP misuse - confidentiality and integrity are each rated Low, availability is unaffected, and the overall CVSS base score is 3.8; no public exploit code has been identified at time of analysis and there is no indication of active exploitation. The exposure is realistic only in deployments that provision narrowly-scoped admin API tokens and treat Admin.Read as a true read-only boundary.
Improper enforcement of mutual TLS client-certificate verification in IBM Financial Transaction Manager (FTM) for Red Hat OpenShift can let an unauthenticated remote attacker who already holds a position on the network path obtain sensitive information from the service. The flaw (CWE-295) is rated CVSS 3.1 3.7 (AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N), meaning exploitation is remotely reachable yet constrained by high attack complexity and yields only limited confidentiality impact with no integrity or availability effect. No public exploit code has been identified at time of analysis and there is no confirmed active exploitation (not in CISA KEV); the overall risk is genuinely low-priority despite remote reachability, and a vendor patch is available.
The GNU C Library dynamic loader (ld.so) in glibc 2.14 through 2.44 mishandles DT_RPATH/DT_RUNPATH values that begin with the literal string $ORIGIN immediately followed by a NUL byte or a '/' character, producing a stack-based buffer overflow (CWE-121) when such a path is parsed in a setuid or setgid (AT_SECURE) program. A local attacker who can execute an affected privileged binary causes the loader to read past the end of the path buffer and write past the end of a stack-allocated internal buffer, resulting in a loader crash (denial of service) and limited disclosure of process memory, with a low CVSS 3.6 score reflecting the local, high-complexity, non-privilege-escalating nature of the flaw. There is no public exploit identified at time of analysis and no CISA KEV listing; exploitation requires local access (PR:L) plus the pre-existing, non-universal condition that a specially-linked setuid/setgid binary exists on the host and is executable by the attacker, which is why all signals converge on this being a LOW real-world priority despite the ubiquity of glibc.
Information disclosure in the Dark Reader browser extension allows an attacker-controlled website to trigger the extension's image-inversion pipeline into fetching an unauthenticated icon-like bitmap from a server on the victim's local network or localhost, leaking limited information about that resource. Affected users are those running Dark Reader in Firefox before 4.9.126, or in other browsers before 4.9.128 - the issue is unauthenticated (CVSS PR:N) but requires the victim to visit a malicious page (UI:R) and depends on a locally reachable server actually listening, with high attack complexity given the need for the resource to sit at a known public-like HTTPS URL and be classified as requiring inversion. Vendor-released patches are available as version 4.9.126 (Firefox) and 4.9.128 (other browsers), and no public exploit code was identified at time of analysis; EPSS-style exploitation likelihood was not provided, and the overall risk is low (CVSS 3.4).
Out-of-bounds read in radare2's Lua 5.3 bytecode function parser affects all versions prior to 6.2.0, allowing a crafted input file to push parser reads past the end of the allocated buffer. Exploitation is local and requires user interaction: an analyst must open or inspect an attacker-supplied Lua 5.3 bytecode file whose function-name string ends exactly at the input-buffer boundary, after which the parser reads two integers and three one-byte metadata fields beyond the allocation, producing invalid parser results or terminating the process. Per the CVSS 3.1 vector (AV:L/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:L) there is no memory disclosure, no code execution, and no privilege escalation - the realistic outcome is a crash of the analyst's own radare2 instance - and there is no public exploit identified at time of analysis and no confirmed active exploitation. The flaw is fixed in radare2 6.2.0.
Opening a crafted Mach-O file in radare2 versions prior to 6.2.0 can trigger a heap out-of-bounds read in the LC_DATA_IN_CODE parser, potentially terminating the process. The flaw is only reachable when the non-default bin.verbose option is enabled and the crafted file declares an LC_DATA_IN_CODE datasize that is not a multiple of the data_in_code_entry structure size, causing the parser's final loop iteration to read past the allocated buffer. Impact is limited to a denial-of-service style crash (CVSS 2.5, AV:L/AC:H/PR:N/UI:R, availability-only); no attacker-observable memory disclosure has been demonstrated, and no public exploit code or confirmed active exploitation was identified at time of analysis.
OpenBao Agent's exec rendering mode leaks secrets to stdout when an error condition occurs, primarily when the configured num_retries is exhausted, exposing credentials supplied via env_template to anyone who can read the agent's output stream or aggregated logs. All OpenBao versions before v2.6.0 are affected; the defect is inherited from the original Vault codebase and the fix is a one-line change that discards the template runner's output stream during failures. The assessed CVSS 3.1 vector (AV:L/AC:H/PR:L/UI:N/S:U/C:H/I:N/A:N) reflects a local, non-default deployment prerequisite - exec rendering mode combined with env_template - and there is no public exploit identified at time of analysis, making this a genuine but low-priority information-disclosure issue.
API token disclosure in the Weblate wlc command-line client before 2.0.1 allows an attacker-controlled project configuration to redirect an unscoped API token to a server the attacker controls. When wlc is run inside an untrusted repository, pull-request checkout, or a directory with an untrusted ancestor whose .weblate, .weblate.ini, or weblate.ini sets an API url, and the operator supplies a token through WLC_KEY or --key without also pinning WLC_URL or --url, the client transmits that token to the project-configured destination. Exploitation requires local execution of wlc, all three configuration preconditions to hold simultaneously, and user interaction; no public exploit was identified at time of analysis, and the overall risk is low (CVSS 2.3 / assessed CVSS:3.1/AV:L/AC:H/PR:N/UI:R/S:C/C:L/I:N/A:N).
Retained OAuth refresh tokens on hex.pm allow a user whose organization membership or session has ended to keep reading that organization's private packages and documentation tarballs. Refresh tokens are issued with the same issuer, audience and repository:<org> scope claims as access tokens, and the CDN that serves private repositories validates only the JWT signature and time claims, performing no membership lookup and unable to tell the two token types apart, so revocation takes effect only when the 30-day refresh token expires instead of at the 30-minute access-token lifetime. The actor must be an authenticated party who was a member when the token was granted, access is read-only and limited to organizations the account belonged to at grant time, and no public exploit has been identified at time of analysis; CVSS 4.0 base score is 2.3 (independent assessment CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N).
SQL injection in TDuckCloud tduck-platform through version 5.3 lets an authenticated user inject arbitrary SQL through the orders[0].column sort parameter of paginated listing endpoints, which MyBatis-Plus concatenates directly into the ORDER BY clause. Successful exploitation primarily enables inferential (blind) reading of database contents, consistent with the assessed CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N profile and the vendor's CVSS 4.0 score of 5.3 with PR:L and uniformly Low impacts. Proof-of-concept exploitation is publicly available (E:P, matching VulDB's 'exploit publicly available' note), but there is no CISA KEV entry, so confirmed active exploitation in the wild has not been established; no EPSS score was supplied for this record.
Code injection in mtrano APENCMS (rolling release, all builds up to commit 6546096d354153309693efabb9a0d824628ed4f5) lets a remote attacker who already holds high-privilege template or site administration access reach a PHP eval() call in cms/weasel.php through the $_CMS['site'] argument and execute arbitrary code on the server. Publicly available exploit code exists, while the vendor was contacted and did not respond, so no fixed build and no published version boundaries are available. The CNA's CVSS 4.0 base score of 1.9 (Low, all impacts rated VC:L/VI:L/VA:L) understates the technical severity: with attacker-controlled input flowing into eval(), successful exploitation is effectively full server-side code execution, and the attacker must already control the template/site variable (PR:H), consistent with template or site administrator rights rather than an unauthenticated default-configuration attack.