Authenticated command injection in reNgine up to 2.2.0 allows a low-privileged user to execute arbitrary OS commands on the server by supplying a crafted target Name, which is passed to a shell command during the subdomain_discovery task. Publicly available exploit code exists (GitHub issue #1561), and the upstream fix PR #1562 awaits acceptance, so no released patched version is confirmed. Independent assessment rates the impact as high (CVSS 3.1: 8.8), while the vendor's CVSS 4.0 score of 2.1 reflects limited scope and low impact ratings, creating a sharp conflict in severity signals; no CISA KEV confirmation was provided.
Stored cross-site scripting in SourceCodester Simple Student Information System 1.0 lets remote, unauthenticated attackers inject arbitrary JavaScript through the firstname and lastname fields processed by /register.php, which executes in the browser of any victim who later views the affected profile page. Publicly available exploit code exists, and the vendor has not released a patch; this is a low-to-moderate priority issue despite the remote attack vector because successful exploitation requires the victim to view the tainted page (UI:P) and impact is limited to client-side script execution with no server-side code execution, no confidentiality loss of stored data, and no availability impact.
Denial of service in vLLM up to and including 0.31.0 lets a remote, authenticated user who can submit inference requests crash the serving engine by combining a prompt_embeds input with token penalty parameters (frequency, presence, or repetition), a combination that drives get_token_bin_counts_and_mask in vllm/model_executor/layers/utils.py into a fault. Publicly available exploit code exists in the form of a published reproduction script, and the vendor had not responded to the early issue report at the time of analysis, so unpatched, internet-exposed or multi-tenant deployments that accept such requests are exposed to availability loss. Impact is limited to availability of the serving process - no data disclosure or integrity compromise has been demonstrated.
An out-of-bounds read in vLLM up to and including version 0.31.0 lets a caller with low-privilege access to the completions API drive the SSM convolution path past the end of a buffer, degrading availability of the inference worker. The flaw is confined to deployments that serve Mamba/Mamba2-architecture models, because it lives in conv_ssm_forward of vllm/model_executor/layers/mamba/mamba_mixer2.py; instances running only standard transformer models are unlikely to reach the vulnerable code. Publicly available exploit code exists, but there is no confirmed active exploitation (not listed in CISA KEV); the independent assessment bounds impact to low availability loss with no confidentiality or integrity effect, making this a moderate rather than top-tier priority.
Cross-site scripting in imgproxy up to 4.0.17 allows remote, unauthenticated attackers to submit crafted SVG content that bypasses the sanitizeElement routine in the SVG handler, potentially executing client-side script when a victim renders the processed image in a browser. Publicly available exploit code exists for the flaw, though no confirmed active exploitation (CISA KEV) is indicated in the provided data. The issue is medium severity because exploitation requires the processed SVG to be rendered in a browser and the deployment to pass SVG through imgproxy's SVG processing path rather than blocking or rasterizing it.
Cross-site scripting (XSS) in SourceCodester Drug Recommendation System 1.0 allows remote attackers to inject malicious scripts via the Admin/add_drug.php endpoint, which execute in an authenticated administrator's browser when the drug data is rendered. The attack requires user interaction and typically administrator-level access to the vulnerable form, and while publicly available exploit code exists, the overall risk is assessed as low-to-moderate due to these prerequisites and the limited browser-context impact.
Improper authorization (CWE-285) in jshERP through version 3.5 allows an authenticated, low-privileged user to manipulate financial receipt records by invoking the updateAccountHeadAndDetail function in AccountHeadService.java, which fails to enforce proper permission checks. The flaw is remotely exploitable with low attack complexity and no user interaction (CVSS 4.0 base 5.3; assessed CVSS 3.1 AV:N/AC:L/PR:L/UI:N), but exploitation requires a valid low-privilege account, so the practical risk is insider abuse or misuse of any obtained credentials rather than unauthenticated external compromise. Impact is limited to integrity and availability of financial receipt data (VI:L/VA:L) with no confidentiality exposure, and publicly available exploit code exists (referenced in GitHub issue 167); the vendor was notified via an issue report but has not responded, and no vendor-released patch was identified at time of analysis.
Authenticated attackers can manipulate the goodsCount parameter in Newbee Mall's shopping cart quantity handler (endpoint /jshERP-boot/accountHead/updateAccountHeadAndDetail) to cause business logic errors, potentially leading to incorrect order totals or inventory records. The vulnerability affects all versions up to 2.7.5, and publicly available exploit code exists; however, active exploitation has not been confirmed (not in CISA KEV). With a CVSS score of 5.3 and impact limited to integrity (no confidentiality or availability loss), the overall risk is medium/low.
An authenticated authorization bypass in PickMall Lilishop 4.2.4 and earlier allows any buyer-level user to read other members' invoice/receipt data by manipulating the memberId parameter on the /buyer/trade/receipt endpoint. Publicly available exploit code exists, and the vendor has not responded to the reported issue, but the vulnerability only exposes limited invoice information (low confidentiality impact) and requires a valid buyer session, so it does not warrant emergency prioritization.
SQL injection in SourceCodester Performance Indicator System 1.0 allows low-privilege authenticated attackers to manipulate the 'Category' parameter in /opils/admin/view_product.php, leading to limited confidentiality, integrity, and availability impact. Publicly available exploit code exists, and the CVSS vector requires PR:L, meaning attackers must first obtain valid low-privilege credentials to the admin interface. There is no evidence of confirmed active exploitation (CISA KEV), and the EPSS score is not available.
SQL injection in girishsaraf's Online Appointment Booking System allows an authenticated low-privilege user to inject arbitrary SQL through the Doctor and appointment parameters handled by book.php, exposing or corrupting booking-related data. All commits up to f427b4757128ca253d33d0cc4e87bbb9c999a4d5 are affected because the project ships as an unversioned rolling release, and the maintainer has not responded to the vendor notification, so no fixed release exists. Exploitation is remote and publicly available exploit code exists (no CISA KEV listing), but the assessed impact is partial rather than full database takeover, making this a genuine but low-to-moderate priority issue.
Cross-site request forgery in SourceCodester Drug Recommendation System 1.0 allows a remote attacker to make an authenticated victim's browser submit unintended state-changing requests, because the application's forms do not validate anti-CSRF tokens (CWE-352). The attacker needs no credentials of their own (CVSS:3.1 AV:N/AC:L/PR:N/UI:R/S:U), but success requires luring a logged-in victim to attacker-controlled content while their session is active, and the resulting impact is limited to low integrity with no confidentiality or availability loss. Publicly available exploit code exists (published in the hackliu/Vulnerability-Reports repository as VULN-06-CSRF-Missing-Tokens.md), but there is no evidence of confirmed active exploitation (not listed in CISA KEV), so this should be treated as a low-to-moderate real-world risk rather than a mass-exploitation event.
SQL injection in the school_year parameter handled by search_class.php allows an authenticated low-privileged user to alter database queries in the KusalkaSilva Learning-Management-System, a PHP-based application distributed on a rolling-release basis with no version identifiers. Publicly available exploit code exists, and the vendor has not responded to the issue report filed against the project repository, so no fixed release can be confirmed. Because exploitation requires a valid low-privileged session with access to the class-search functionality (CVSS 4.0 vector PR:L, UI:N, AV:N), anonymous attackers cannot reach the vulnerable parameter, which limits exposure; the assessed impact is low confidentiality, integrity, and availability loss.
Reflected cross-site scripting in SourceCodester Simple Student Information System 1.0 allows a low-privileged authenticated attacker to inject script through the 'searchbox' argument handled by the clean function in searchresults.php, which reflects the value into the results page without proper output encoding. The CVSS 4.0 vector (AV:N/AC:L/AT:N/PR:L/UI:P) shows the attacker needs low-level authentication to reach the search page, while successful execution requires a logged-in victim to view the attacker-influenced results page, so impact is confined to integrity within the victim's browser session (VC:N/VI:L/VA:N). Publicly available exploit code exists (referenced from a public GitHub issue), but there is no CISA KEV listing indicating confirmed active exploitation; combined with an assessed CVSS 3.1 of 5.1, this is a low-to-moderate priority issue rather than an urgent one.
Improper authorization in SpringBlade up to 5.0.1 allows an authenticated remote user to access role or user detail records they do not own by manipulating the ID parameter in the RoleController User Detail endpoint. The flaw is a CWE-285 improper authorization issue (insecure direct object reference) that affects only confidentiality, with no integrity or availability impact and no code execution; publicly available exploit code exists, and no CISA KEV listing is present. The independent assessment rates this as a moderate, not urgent, priority, and notes that while the CVSS vector indicates high privileges (PR:H), the practical privilege level may be lower because any logged-in user could likely exploit the IDOR.
Unauthorized password changes are possible in PHPGurukul User Registration & Login and User Management System 3.3 because the Change Password Handler at loginsystem/admin/change-password.php fails to properly validate the 'currentpassword' parameter, an incorrect authorization flaw (CWE-863) that lets a caller manipulate the request and bypass the intended current-password check. Per the supplied CVSS 4.0 vector (PR:H), the attacker must already hold an authenticated session with privileges, so this is not reachable by unauthenticated remote users, and the assessed impact is limited to low confidentiality, integrity, and availability within a single security scope. Publicly available exploit code exists (a GitHub proof-of-concept repository is referenced), making the technique easy to reproduce, but no CISA KEV entry or EPSS score was provided, so confirmed active exploitation is not established by the available data; overall this is a low-priority issue despite being network-reachable.
Improper authorization in the SpringBlade Java microservice framework (all versions up to 5.0.1) lets an authenticated user holding high privileges manipulate the initPassword argument exposed by the Parameter Submit Management controller (ParamController) and change system parameters they should not be able to control. The endpoint is remotely reachable, but exploitation requires an existing elevated session with no user interaction, and the assessed impact is LOW across confidentiality, integrity and availability (CVSS 3.1 AV:N/AC:L/PR:H/UI:N/S:U/C:L/I:L/A:L). Publicly available exploit code exists and the project has not responded to the issue report, so no vendor-released patch is available at time of analysis.
Uncontrolled recursion in yawkat LZ4 Java's net.jpountz.lz4.LZ4BlockInputStream can crash the decoding thread: when the stream is explicitly configured with stopOnEmptyBlock set to false, each well-formed empty LZ4 block causes refill() to call itself recursively, so an attacker-supplied compressed stream containing a long run of empty blocks drives the stack to exhaustion and throws StackOverflowError. Only applications that both use LZ4BlockInputStream with the non-default stopOnEmptyBlock=false setting and feed it attacker-controlled or injectable compressed data are exposed; the default stopOnEmptyBlock=true is not affected, no authentication or user interaction is required under those conditions (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:N/A:L), and impact is limited to availability of the decoding thread with no memory corruption, confidentiality, or integrity effect. There is no public exploit identified at time of analysis and the flaw is fixed in version 1.11.4.
The sshd and ssh components of OpenSSH before 10.6 can be made to negotiate and use the LZ77 dictionary-coder compression scheme, a behavior the 'Crossing the Streams' research (arXiv 2609.07709) identifies as contraindicated because compression-derived information leakage can be measured from an observed session. A remote, unauthenticated attacker who can observe and influence the target connection and repeatedly sample it may recover a limited amount of confidential data, but only where the vulnerable compression mode is actually in use; integrity and availability are unaffected. This is a low-severity, low-priority issue (CVSS 3.1 base 3.7; AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N) despite being network-reachable and unauthenticated, because exploitation demands an active compression configuration, a privileged observation vantage point, and sustained measurement - no public exploit code has been identified at time of analysis and there is no confirmed active exploitation (CISA KEV).
URL injection and server-side request forgery (SSRF) in i18next-http-backend prior to 4.0.2 allows attacker-controlled language or namespace values to redirect translation resource requests away from the intended origin, but only when the application configures a custom loadPath or addPath template that begins directly with the {{lng}} or {{ns}} placeholder. Because the placeholder must occupy the URL's structural beginning for a colon to form an absolute URL (or, in browsers, a double-slash namespace to form a protocol-relative URL), the default '/locales/{{lng}}/{{ns}}.json' template and any template with a leading path or origin prefix are unaffected; this uncommon configuration and the need to control the language/namespace input drive the High attack complexity rating. Impact is limited to low confidentiality loss (blind SSRF/URL injection with no integrity or availability effect), and all signals indicate a low-priority issue - no public exploit code and no confirmed active exploitation were identified at time of analysis. The issue is fixed in version 4.0.2.
Improper input validation (CWE-20) in HCL BigFix Service Management lets an attacker who already holds a low-privilege authenticated account on the adjacent network submit unexpected or malformed data, producing processing errors, limited business-logic bypasses, and unintended application behavior. The assessed vector (CVSS:3.1/AV:A/AC:L/PR:L/UI:N/S:U/C:N/I:L/A:N, score 3.5) confines impact to a low-integrity effect with no confidentiality or availability loss, and exploitation requires both adjacency to the deployment and an existing account, so this is a low-severity, low-priority item despite being a published CVE. No public exploit code or CISA KEV listing was identified at time of analysis, and no EPSS score was supplied; HCL's own tag of 'Authentication Bypass' is broader than the assessed integrity-only impact and should not be read as a full authentication defeat.
Ubuntu Linux carries CVE-2026-55453, a security issue that has so far been reported only by the vendor (Ubuntu) with no published description, no CVSS score or vector, no CWE classification, and no CPE or version data. Because the flaw is entirely undisclosed, it is not possible to state which package, subsystem, or Ubuntu release is affected, what an attacker would need to reach it, or what the resulting impact would be. There is no public exploit identified at time of analysis, no KEV listing, no EPSS score, and no confirmed patch availability, so the item should be treated as an unassessed placeholder pending the corresponding Ubuntu Security Notice (USN) and NVD entry.
An out-of-bounds read in Google Chrome's Skia graphics rendering library (CWE-125) allows an attacker to read a limited amount of process memory when a victim is socially engineered into opening a crafted file that Chrome renders, with the flaw present in all Chrome builds prior to 155.0.8059.39. The assessed vector (CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:L/I:N/A:N, base 3.3) shows no authentication requirement but a mandatory victim interaction and only a low-confidentiality memory exposure with no integrity or availability impact, which keeps real-world risk small; Google rates it Low. EPSS is 0.11% (1st percentile), no public exploit has been identified at time of analysis, and Chrome builds at or above 155.0.8059.39 are not affected.
Misinterpretation of the literal configuration value "none" in sshd in OpenSSH before 10.6 causes the daemon to treat the keyword as a filename instead of as a feature-disable value, yielding limited integrity and availability impact with no confidentiality loss. Per our independent assessment this is a low-severity, low-priority issue despite being in ubiquitous software: exploitation is local only (AV:L), requires high attack complexity (AC:H) and high privileges (PR:H), and depends entirely on an administrator having set an sshd option to the literal string "none" in a non-default configuration, so there is no remote or default-configuration exposure. Exploitation status: no public exploit identified at time of analysis, and the issue is not listed in CISA KEV; the only reference is the upstream OpenSSH 10.6 release notes.
ssh-keygen in OpenSSH versions before 10.6 can assign incorrect expiration timestamps to generated certificates when a Daylight Saving Time transition is mishandled, producing a bounded integrity error in the certificate's validity window. The issue is local: it requires a user to run ssh-keygen for certificate creation or signing, and it only manifests around DST changes, with slightly more pronounced effects noted for certain Antarctic time zones. With a CVSS score of 2.5 and no public exploit identified at time of analysis, this is a low-priority correctness bug rather than a remotely exploitable security flaw.
Cross-origin data disclosure in Google Chrome versions prior to 155.0.8059.39 stems from a use-after-release defect in the browser's Core component, but it is not independently exploitable: a remote attacker must first have already compromised Chrome's sandboxed renderer process through a separate, unrelated vulnerability, then persuade the victim to load a crafted HTML page. Only limited confidentiality impact is possible (CVSS 3.1, AV:N/AC:H/PR:N/UI:R/S:U/C:L/I:N/A:N; Chromium security severity: Low), with high attack complexity reflecting the non-deterministic timing needed to touch the freed resource. No public exploit code has been identified at time of analysis and there is no confirmed active exploitation (CISA SSVC exploitation: none), so this is best treated as a chain-stage bug rather than a standalone threat.
Cross-origin information disclosure in Google Chrome prior to 155.0.8059.39 lets a remote attacker leak a limited amount of data from other origins when a victim loads a crafted HTML page that abuses an observable rendering discrepancy in Chrome's Animation implementation. Exploitation requires user interaction (the victim must open the attacker's page) and is rated high complexity (AV:N/AC:H/PR:N/UI:R/S:U/C:L/I:N/A:N, CVSS 3.1), reflecting repeated timing/rendering measurement rather than a straightforward one-shot extraction; there is no code execution, data modification, or denial of service, and no public exploit identified at time of analysis. CISA's SSVC assessment rates exploitation as 'none' and automation as 'no', and no KEV listing applies, so this is a low-priority, low-real-risk issue despite affecting a ubiquitous browser.
Cross-origin information disclosure in Google Chrome (desktop stable channel) affects releases prior to 155.0.8059.39, where an observable discrepancy in the browser's Scroll behavior can be measured to infer data from other origins. Exploitation is not possible simply by visiting a web page: the attacker must first get the victim to install a crafted Chrome extension, which is the user-interaction requirement (UI:R) and the main driver of the high attack-complexity rating (AC:H) in the assessed vector CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:L/I:N/A:N. No public exploit code has been identified at time of analysis, CISA's SSVC assessment records Exploitation as 'none' and Automatable as 'no' with partial technical impact, and the impact is confined to a low-confidentiality leak with no integrity or availability effect. This is a genuinely low-priority issue despite affecting a ubiquitous product, and the practical exposure window is further narrowed by Chrome's automatic update mechanism.
Missing authorization in Google Chrome's Google Lens component (versions prior to 155.0.8059.39) lets a remote, unauthenticated attacker read cross-origin data when a victim loads a crafted HTML page - but only after that attacker has already compromised the browser's renderer process, which is the gating precondition and makes this a chained, secondary flaw rather than a standalone entry point. Exploitation additionally requires user interaction (UI:R) and yields only low-severity confidentiality impact with no integrity or availability effect, so the assessed vector is CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:L/I:N/A:N for a base score of 3.1, rated Medium by Chromium. Vendor-released patch: 155.0.8059.39; no public exploit identified at time of analysis, and no CISA KEV status or EPSS signal indicates active exploitation.
OpenSSH sshd builds compiled against the macOS 27 (or later) SDK silently lose their sandbox, removing a containment layer that normally limits the damage an attacker can do if they manage to compromise or corrupt the sshd process. Only OpenSSH through 10.6 built against that specific SDK is affected, which is far from the default configuration for Linux, BSD, or older-macOS deployments. This is a low-severity defense-in-depth regression rather than a directly exploitable flaw - the sandbox loss causes no impact on its own and only produces harm when chained with a separate exploitable sshd vulnerability that the sandbox would otherwise have contained; no public exploit was identified at time of analysis, and the CVSS 3.1 score (AV:N/AC:H/PR:L/UI:N/S:U/C:N/I:L/A:N) reflects that chaining requirement and high attack complexity.
On iOS builds of Quasar Framework (all versions prior to 2.32.2), the openURL() utility can be tricked into invoking attacker-controlled page content instead of the native SafariViewController bridge, breaking external navigation, login redirects, and payment redirects with a thrown TypeError. The trigger requires an app built for iOS (Cordova/Capacitor where Platform.is.ios is true) plus a QEditor, QSelect, or QChatMessage configured to render untrusted HTML that can define an element named 'SafariViewController'; a legitimate user must then attempt a URL-opening action. Impact is limited to denial of the URL-opening function (CVSS 3.1, AV:N/AC:H/PR:N/UI:R, C:N/I:N/A:L) and the bug is fixed in 2.32.2; no public exploit code or confirmed active exploitation was identified at time of analysis.
Microsoft Kiota, the OpenAPI-based HTTP client code generator, fails to validate the x-ai-capabilities.response_semantics.oauth_card_path value before copying it into a generated API plugin manifest, allowing path-traversal or absolute-URI references to survive into the manifest in versions 1.25.1 through 1.34.x. An attacker who controls or compromises the OpenAPI description can inject a malicious oauth_card_path (for example '../../../../etc/passwd', 'C:\\Windows\\...', a UNC path, or 'file:///etc/passwd') so that a downstream host resolving the plugin manifest crosses the intended plugin-package boundary or loads an unintended authentication card. The attacker vector is network-based and unauthenticated (PR:N), but exploitation requires a developer to actively run Kiota generation against the malicious description and a separate downstream consumer to resolve the reference; Kiota performs no file read or code execution during generation itself, so impact is limited to low integrity (I:L) and there is no public exploit identified at time of analysis. The issue is fixed in version 1.35.0.
This is a genuinely low-priority issue by every available signal, not a high-CVSS alarm. Local privilege retention in OpenSSH sshd-session on obsolete platforms (QNX 6, SCO OpenServer 5) allows an authenticated user to cause sshd-session to run with root privileges when GatewayPorts or StreamLocalForwarding is enabled and the host lacks support for file-descriptor passing and unprivileged PTY allocation. Exploitation is confined to those specific configurations, no public exploit has been identified at time of analysis, and the assessed CVSS is 2.9.
Stored cross-site scripting in the Class2HTML report generator shipped with Apache Commons BCEL before 6.13.0 lets attacker-controlled strings from a Java class file (class, field and method names, constant-pool entries, attribute values and anchors) be written into generated HTML reports without escaping, so script executes in the browser of anyone who opens the report. The issue is unauthenticated (CVSS 4.0 PR:N) but highly conditional: the deployment must actually use the Class2HTML/AttributeHTML/CodeHTML/ConstantHTML utilities to render HTML from class files an attacker can influence, and a victim must then open the resulting report (UI:P), which is why the CVSS 4.0 base score is only 2.3 because attack requirements are present (AT:P) even though the independent assessment rates the same flaw 7.1 under CVSS 3.1 with scope change (S:C). This is not reachable through ordinary BCEL bytecode parsing or manipulation; no public exploit has been identified at time of analysis and the flaw is fixed in 6.13.0.
Memory disclosure in Arista access points configured with VXLAN tunnelling and L2-proxy (the VESPA use-case) lets an unauthenticated adjacent-network wireless client associated to a tunnelled SSID send a crafted packet and cause the AP to expose memory contents in network traffic. The flaw is an out-of-bounds read (CWE-125) rated CVSS 4.0 2.3, with the independent assessment bounding impact to low confidentiality loss only; no write primitive, integrity impact, availability impact, or remote code execution is possible. The issue is low priority despite alarmist RCE/buffer-overflow tags, and no public exploit was identified at time of analysis.
RediSearch query injection in @langchain/redis (the Redis vector store provider for LangChain.js) prior to version 1.1.1 lets a low-privileged attacker who can influence the values passed into structured RediSearch TAG or TEXT filters inject unescaped RediSearch syntax and thereby broaden or otherwise alter the generated search query. Where an application uses such a filter as a tenant or document-access boundary, the manipulated query can return indexed documents outside the attacker's intended scope, resulting in limited confidentiality impact (read/exposure only, no write, code execution, or availability effect). Exploitation is highly conditional: the assessed vector is CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N (CVSS 4.0 base 2.3), no public exploit code has been identified, and no confirmed active exploitation (CISA KEV) or EPSS signal was provided, so overall risk is low; the fix is available in @langchain/redis 1.1.1.
Heap corruption in GNU Aspell's personal-wordlist handling lets a local attacker crash a user's spell-checking session by supplying a crafted dictionary file. All Aspell releases before 0.60.8.3 are affected: when a wordlist containing a word whose length is an exact multiple of 256 bytes is loaded, WritableDict::add() in modules/speller/default/writable.cpp stores the word length in a single byte, truncating it and corrupting heap memory, which results in denial of service rather than code execution or data disclosure. No public exploit identified at time of analysis, and the issue is not listed in CISA KEV; the vendor fix was published as commit 782ce94e4dc71eaec4ee1bd945eb3b9c47c5387d and is slated for release in version 0.60.8.3. With a CVSS 4.0 base score of 1.8 and an independent assessment of CVSS:3.1/AV:L/AC:H/PR:N/UI:R/S:U/C:N/I:N/A:L, this is a genuinely low-priority issue: the attack is local-only, needs the victim to actively load an attacker-supplied dictionary, requires the specific multiple-of-256 length condition, and yields availability loss only.
Heap out-of-bounds read in GNU Aspell's binary dictionary loader (ReadOnlyDict::load() in readonly_ws.cpp) lets a crafted .rws dictionary file trigger limited heap memory disclosure or a crash of the aspell process. The victim must deliberately run aspell against the attacker-supplied dictionary via --master, --dict-dir, or an equivalent configuration setting, so the vector is local, high-complexity and dependent on user interaction, and no privileges are required (PR:N). Impact is confined to limited information disclosure and denial of service with no code execution or integrity impact, and no public exploit has been identified at time of analysis.
GNU Aspell's prezip-bin decompressor suffers a heap-based buffer overflow (CWE-122) that lets a crafted prezip-compressed file corrupt heap memory and crash the process. The issue affects Aspell versions prior to 0.60.8.3 and is only reachable through a local chain: an attacker must deliver a malicious compressed file and then persuade a user to run prezip-bin against it, so no privileges are required (PR:N) but user interaction and high attack complexity apply. The assessed impact is availability-only (VC:N/VI:N/VA:L) with a low CVSS 4.0 base score of 1.8 and an equivalent CVSS 3.1 assessment of AV:L/AC:H/PR:N/UI:R/S:U/C:N/I:N/A:L; no public exploit code has been identified at time of analysis and there is no CISA KEV listing. Multiple signals - trivial score, local-only vector, interaction requirement and crash-only outcome - converge on a genuinely low-priority issue.