CVE Database

9Router is an AI router & token saver. Prior to 0.5.2, 9router validates image URLs by resolving the host before fetching, but open-sse/translator/concerns/image.js performs the later server-side image fetch with a separate DNS resolution. An authenticated attacker with access to the LLM proxy can use a vision-capable model and an attacker-controlled DNS name that first resolves to a public IP and then rebinds to an internal address, allowing server-side requests to internal-only HTTP services. This issue is fixed in version 0.5.2.

9Router is an AI router & token saver. Prior to 0.4.80, the dashboard login rate limiter in src/lib/auth/loginLimiter.js derives the client identity from the attacker-controlled X-Forwarded-For HTTP header, and src/app/api/auth/login/route.js uses that spoofable value for checkLock and recordFail. A remote attacker can rotate the X-Forwarded-For value on each login attempt to receive a fresh rate-limit bucket, bypass the 5-attempt threshold and progressive lockout durations, and perform unlimited brute-force attempts against the dashboard password. This issue is fixed in version 0.4.80.

MaxKB is an open-source AI assistant for enterprise. Prior to 2.10.0-lts, MaxKB tool import functionality in apps/tools/serializers/tool.py and MCP referencing mode in apps/application/chat_pipeline/step/chat_step/impl/base_chat_step.py do not consistently validate MCP transport type, allowing an authenticated user to import a .tool file containing stdio transport with malicious commands and trigger the configuration through an AI Chat node so MultiServerMCPClient executes arbitrary system commands. This issue is fixed in version 2.10.0-lts.

osquery is a SQL powered operating system instrumentation, monitoring, and analytics framework. Prior to 5.23.1, on Windows, a local unprivileged attacker can cause a heap buffer out-of-bounds write if there is a query of the authenticode table targeting a maliciously crafted binary, due to publisher information parsing in getOriginalProgramName. If exploited successfully, this could allow a potential local privilege escalation from standard user to SYSTEM. This issue is fixed in version 5.23.1.

osquery is a SQL powered operating system instrumentation, monitoring, and analytics framework. Prior to 5.23.1, on Windows, a local unprivileged attacker can cause a heap buffer out-of-bounds write if there is a query of the processes table targeting a maliciously crafted process, due to unchecked PEB string lengths in process command-line and current-directory reads. If exploited successfully, this could allow a potential local privilege escalation from standard user to SYSTEM. This issue is fixed in version 5.23.1.

Several Grafana API endpoints, some of them unauthenticated, do not limit the size of the request body before processing it. An attacker can send very large payloads that force excessive memory allocation, potentially exhausting memory and causing a denial of service.

PraisonAI (pip package praisonaiagents) before 1.6.78 contains an unsafe dynamic module loading vulnerability in AgentFlow._resolve_pydantic_class (src/praisonai-agents/praisonaiagents/workflows/workflows.py). When a workflow step uses a string output_pydantic reference, the framework locates and imports a sibling tools.py from the workflow file's directory via importlib exec_module without sandboxing, ignoring the PRAISONAI_ALLOW_*_TOOLS environment variables. An attacker who controls a workflow file and its sibling tools.py can execute arbitrary Python code with the workflow runner's privileges when the workflow is executed via WorkflowManager or after load_yaml.

PraisonAI versions before 4.6.78 contain an allowlist bypass vulnerability in shell command execution that allows attackers to execute restricted commands via find's built-in -exec, -execdir, and -delete actions. Attackers can craft find commands with these built-in actions to read blocked files, delete files, or execute non-allowlisted binaries without triggering shell metacharacter filters.

PraisonAI before 4.6.78 contains an unauthenticated server-side request forgery vulnerability in the Jobs API /api/v1/runs endpoint. The webhook_url parameter is validated at request time but re-resolved at connection time, allowing attackers to use DNS rebinding to reach internal services with a blind SSRF attack.

In JetBrains TeamCity before 2026.1.2 pipeline modification was possible due to improper permission checks

In JetBrains TeamCity before 2026.1.2 stored XSS via unauthenticated agent registration was possible

In JetBrains TeamCity before 2026.1.2 stored XSS on the cloud profile page was possible via agent-reported data

In JetBrains TeamCity before 2026.1.2 arbitrary file access was possible via the Perforce VCS integration

Capgo before 12.128.2 contains an authentication bypass vulnerability in the password change endpoint that allows attackers to change user passwords without requiring current password confirmation. Attackers with temporary session access can exploit this flaw to permanently lock out legitimate users and achieve full account takeover.

Capgo before 12.128.2 contains an information disclosure vulnerability in the get_orgs_v7(userid) RPC function that remains publicly invokable despite intended private access controls. Unauthenticated attackers can supply arbitrary user UUIDs to retrieve foreign users' organization membership, roles, management emails, and billing metadata.

Crawl4AI before 0.8.7 contains a server-side request forgery (SSRF) vulnerability in the Docker API server's /crawl/job and /llm/job endpoints, which accept webhook URLs without destination validation. An attacker can supply webhook URLs pointing to private or internal IP ranges, Docker networks, or cloud metadata endpoints (e.g. 169.254.169.254), causing the server to make requests to internal services and potentially expose cloud metadata.

In @capgo/capacitor-updater (Cap-go/capgo) before 12.128.2, the end-to-end encryption scheme distributes the private key to each device that downloads the app. Because the public key can be derived from the private key, an attacker performing a man-in-the-middle attack or compromising the Capgo server can create a validly signed update bundle and cause devices to install an update not produced by the original app maker.

The iDirect iQ200 exposes the /api/identity and /api/ REST API endpoints without authentication. An unauthenticated attacker with network access can retrieve sensitive device information including the serial number, Device ID (DID), Terminal Private Key identifier (TPK), MAC address, and exact firmware version. The DID and TPK are used for satellite network authentication in the iDirect platform, potentially enabling terminal impersonation and network reconnaissance.

The iDirect iQ200 does not validate CSRF tokens on state-changing API endpoints after authentication. The /api/reboot endpoint accepts POST requests authenticated solely by a session cookie that lacks the SameSite attribute. A remote attacker can host a malicious web page that, when visited by an authenticated administrator, automatically submits a cross-site POST request causing an immediate device reboot and satellite link loss. Repeated attacks can sustain a denial-of-service condition.

Lucee CFML Server versions across the 5.3.x, 6.1.x, 6.2.x, and 7.0.x release lines contain a reflected cross-site scripting vulnerability in URL path parsing that allows unauthenticated remote attackers to execute arbitrary JavaScript in a victim's browser by embedding HTML or JavaScript payloads within the request path. Attackers can craft a malicious URL containing injected script content that is reflected in the server's response without proper output encoding, enabling session hijacking or unauthorized actions against the Lucee administrative interface when a victim visits the crafted link.

FlaskBB through 2.2.0, fixed in commit a5da9a5, contains a logic flaw vulnerability that allows authenticated administrators to delete all built-in authorization groups by exploiting a type mismatch in the bulk delete protection check. The bulk AJAX endpoint in the management views compares received JSON integer group IDs against string literals, causing the protection check to always pass, which allows deletion of all six built-in groups and destroys the forum's permission model, potentially rendering the site unusable.

FlaskBB through 2.2.0, fixed in commit acc88cf, contains an authorization bypass vulnerability that allows authenticated moderators to perform unauthorized actions on topics in forums they do not control by submitting crafted topic ID lists. Attackers can include a low-ID topic from a permitted forum as an anchor in a batch request, causing the permission check applied only to the first result to pass, and then execute lock, unlock, delete, or hide actions against topics in unmoderated forums.

Dell Unisphere for PowerMax, version(s) 10.3.0.5 and prior, contain(s) a Deserialization of Untrusted Data vulnerability. A low privileged attacker with remote access could potentially exploit this vulnerability, leading to arbitrary command execution with root privileges.

Dell PowerFlex Manager, Version prior to 5.1.0.1, contain(s) an Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability. A low privileged attacker with remote access could potentially exploit this vulnerability, leading to Information disclosure, Information exposure, and Unauthorized access.

Dell PowerFlex Manager, Version prior to 5.1.0.1, contain(s) an Improper Neutralization of Special Elements used in an SQL Command ('SQL Injection') vulnerability. A low privileged attacker with remote access could potentially exploit this vulnerability, leading to Information exposure.

R-SOFT DMS stores superadmin credentials using a non-salted nested MD5 hash. This allows an attacker who obtain password hash to decode superadmin credentials. Critically, this password cannot be changed except by modifying the configuration file. This issue was fixed in version v3.17-2000.

R-SOFT DMS is vulnerable to Insecure Direct Object Reference (IDOR) attack in multiple file download endpoints. The application fetches files from the database by ID and serves them to whoever requests them, relying only on session authentication, meaning any valid user can access any file. This issue was fixed in version v3.19-2862 and v3.17-2580.

R-SOFT DMS is vulnerable to OS Command Injection in konwertujAction() function. The document converter executes shell commands using unsanitized file paths and format parameters. This allows an authenticated attacker to execute arbitrary system commands with the privileges of the web server user. This issue was fixed in version v3.19-2752 and v3.17-2580.

Out-of-bounds Read, Improper Input Validation vulnerability in Apache IoTDB C++ client. Out-of-bounds reads in IoTDB C++ client TsBlock deserializer crash client process on malformed server data. This issue affects Apache IoTDB C++ client: from 1.3.5 before 1.3.8, from 2.0.5 before 2.0.10. Users are recommended to upgrade to version 2.0.10, which fixes the issue.

Incorrect Authorization, Improper Access Control vulnerability in Apache IoTDB. Authorization bypass in /rest/v2/fastLastQuery exposes last-value data to unauthorized authenticated users. This issue affects Apache IoTDB: from 1.3.5 before 1.3.8, from 2.0.5 before 2.0.10. Users are recommended to upgrade to version 2.0.10, which fixes the issue.