Updated 3 h ago

Vulnerability database

Every flaw the U.S. government’s cybersecurity agency (CISA) has confirmed attackers are using, plus the new WordPress plugin and theme bugs published each day, sorted by what they mean for a small business. We refresh it twice a day. If you see something your website or office runs, tell us.

1,764exploited flaws on CISA’s list

35added in the last 30 days

479new WordPress flaws this week

390,525entries in the database

3,489 results · page 2 of 70

Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57144 · Other software · MervinPraison · reported by GitHubHigh 8.8
PraisonAI is a multi-agent teams system. From version 4.5.110 to before version 4.6.61, praisonai.sandbox.SandlockSandbox is documented and implemented as the kernel-enforced sandbox backend for untrusted code. Its SandboxConfig.native() path lets callers configure allowed filesystem paths and network=False. On systems where the optional sandlock module imports but reports that Landlock is unavailable, SandlockSandbox.execute() and run_command() do not fail closed. They silently fall back to SubprocessSandbox(self.config). That fallback keeps the same high-level native policy object but does not enforce the native filesystem or network boundary during code execution. A sandboxed payload can read files outside the configured allowed path and open network connections despite network=False. This issue has been patched in version 4.6.61.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57142 · Other software · MervinPraison · reported by GitHubHigh 7.8
PraisonAI is a multi-agent teams system. From version 4.5.87 to before version 4.6.61, PraisonAI recipe execution has a dangerous-tool policy that is supposed to block default-denied tools unless the caller explicitly passes allow_dangerous_tools=True. That policy only checks tools declared in TEMPLATE.yaml requires.tools. For steps-based recipes, the actual execution path loads workflow.yaml with YAMLWorkflowParser. That parser resolves agent-level tools: declarations and preserves top-level approve:. Workflow.start() then installs those YAML-approved tools into the approval context. As a result, an untrusted recipe can omit execute_command from TEMPLATE.yaml requires.tools, declare it in workflow.yaml agents.*.tools, and add top-level approve: [execute_command]. The caller did not set allow_dangerous_tools=True, but the recipe policy allows the recipe and the workflow approval path self-approves the critical shell tool. This issue has been patched in version 4.6.61.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57121 · Other software · MervinPraison · reported by GitHubHigh 8.1
PraisonAI is a multi-agent teams system. From version 0.1.4 to before version 0.1.5 in praisonai-platform, a workspace member can permanently delete any resource — projects, agents, issues, labels, issue dependencies, and issue-label attachments — created by the workspace owner or other members. All six content DELETE endpoints enforce workspace membership but perform no ownership or role check. A single malicious or compromised member account can wipe an entire workspace's content irreversibly. This issue has been patched in version 0.1.5.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57117 · Other software · MervinPraison · reported by GitHubHigh 8.8
PraisonAI is a multi-agent teams system. From version 4.6.10 to before version 4.6.59, LocalManagedAgent / SandboxedAgent compute bridging wraps read_file, list_files, and write_file when a compute provider is attached. The bridge converts those file operations into shell command strings using raw path arguments, then sends those strings to shell-backed compute providers. An attacker who can influence a file-tool path argument can break out of the quoted path and execute arbitrary shell commands in the compute environment. With compute="local", commands execute through the local subprocess compute provider on the host. With Docker, commands execute in the container. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57116 · Other software · MervinPraison · reported by GitHubCritical 9.8
PraisonAI is a multi-agent teams system. From version 4.2.1 to before version 4.6.59, PraisonAI's AgentOS FastAPI deployment surface remains unauthenticated in current main and in releases after the published patched version for GHSA-pm96-6xpr-978x / CVE-2026-40151. The chat route directly calls agent.chat(request.message). No-auth and wrong-bearer requests both execute the deployed agent. This is broader than passive metadata disclosure. In any deployment where AgentOS wraps agents with tools, private context, memory, API integrations, or cost-bearing model calls, an unauthenticated reachable client can drive those agents. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57114 · Other software · MervinPraison · reported by GitHubHigh 7.2
PraisonAI is a multi-agent teams system. From version 4.5.126 to before version 4.6.59, PraisonAI's Async Jobs API validates webhook_url when a job request is parsed and again when the internal Job object is constructed. That validation blocks direct loopback/private targets, but it is not bound to the later network request. When a job completes, _send_webhook() passes the original hostname to httpx.AsyncClient.post() with no send-time validation, IP pinning, or guarded transport. An attacker-controlled hostname can therefore resolve to a public IP during Pydantic validation and later resolve to loopback/private/cloud-metadata infrastructure during webhook delivery. This bypasses the intended SSRF guard in current supported releases. This appears to be an incomplete fix / patch bypass for GHSA-8frj-8q3m-xhgm. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-57113 · Other software · MervinPraison · reported by GitHubHigh 8.1
PraisonAI is a multi-agent teams system. From version 2.6.0 to before version 4.6.59, PraisonAI's template loader accepts GitHub template URIs with refs, for example github:owner/repo/template@v1.0.0. The resolver stores the user-controlled template path and ref verbatim, and the cache layer later joins those values into ~/.praison/cache/templates/github/<owner>/<repo>/<template>/<ref> without normalizing each segment or checking that the final path remains inside the template cache root. A crafted ref such as ../../../../../../outside-delete-target therefore escapes the cache directory. The first load can write .cache_meta.json outside the cache. If the normal cache hierarchy for the same owner/repo/template has already been created, the same path reaches shutil.rmtree(cache_path) and removes an attacker-selected outside directory before replacing it with cache metadata. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026Agentfront mcp-from-openapi from 2.2.0 before 2.9.0 contains a server-side request forgery vulnerability that allows attackers to bypass the ssrf.blockedHosts dCVE-2026-115751 · Other software · agentfront · reported by VulnCheckMedium 6.3
Agentfront mcp-from-openapi from 2.2.0 before 2.9.0 contains a server-side request forgery vulnerability that allows attackers to bypass the ssrf.blockedHosts denylist using trailing-dot hostnames. Attackers who control an OpenAPI spec URL or redirect target can append a root dot to a blocked hostname to fetch content from operator-blocked hosts.
Product
mcp-from-openapi
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026mcp-from-openapi from 2.5.0 before 2.9.0 contains a credential exposure vulnerability that allows attackers to capture Authorization and Cookie headers because CVE-2026-115750 · Other software · agentfront · reported by VulnCheckMedium 6.3
mcp-from-openapi from 2.5.0 before 2.9.0 contains a credential exposure vulnerability that allows attackers to capture Authorization and Cookie headers because safeFetch() forwards them on cross-origin redirects. Attackers controlling a spec host or abusing an open redirect can redirect fromURL() spec requests to their own origin to steal bearer tokens or cookies.
Product
mcp-from-openapi
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56840 · Other software · MervinPraison · reported by GitHubHigh 8.8
PraisonAI is a multi-agent teams system. From version 4.5.2 to before version 4.6.59, praisonai.bots.HTTPApproval renders pending tool approval arguments directly into the approval dashboard HTML. An attacker-controlled tool argument can inject JavaScript into that page. When a human opens the approval URL to inspect the risky tool request, the script runs in the dashboard origin and can POST to the same request's /approve/{request_id}/decide endpoint, causing HTTPApproval to return approved=True. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56838 · Other software · MervinPraison · reported by GitHubHigh 7.8
PraisonAI is a multi-agent teams system. From version 4.5.87 to before version 4.6.59, PraisonAI recipe execution blocks default-denied dangerous tools unless the caller explicitly passes allow_dangerous_tools=True. The normal recipe.run() path enforces this with _check_tool_policy(). The streaming path, recipe.run_stream(), loads the same recipe, checks dependencies, and then calls _execute_recipe() without running the dangerous-tool policy check. As a result, a recipe that honestly declares execute_command in TEMPLATE.yaml requires.tools is denied by recipe.run(), but reaches the execution engine through recipe.run_stream() with allow_dangerous_tools=False. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56837 · Other software · MervinPraison · reported by GitHubHigh 8.6
PraisonAI is a multi-agent teams system. From version 4.6.56 to before version 4.6.59, PraisonAI's LinearBot starts a public webhook listener on 0.0.0.0 and treats LINEAR_WEBHOOK_SECRET as optional. When the secret is absent, startup only logs a warning and _handle_webhook() skips Linear-Signature verification entirely. An unauthenticated network caller who can reach the webhook endpoint can submit a forged Linear-Event: AgentSession request. The forged request is parsed, scheduled for background processing, dispatched to _handle_agent_session(), and passed into BotSessionManager.chat(). The bot then attempts to post the agent response back to Linear under the configured bot token. The PoV is offline and deterministic. It does not contact Linear. It calls the webhook handler directly, monkey-patches the outbound Linear comment path, and proves both sides of the boundary. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56836 · Other software · MervinPraison · reported by GitHubHigh 8.2
PraisonAI is a multi-agent teams system. From version 4.5.112 to before version 4.6.59, PraisonAI's installed console entrypoint is Typer-first. In current releases, the recipe command is registered in the Typer app and praisonai recipe serve dispatches to the deprecated Typer command in src/praisonai/praisonai/cli/commands/recipe.py. That Typer command can start the Recipe HTTP server on a non-localhost interface with no authentication. It prints a deprecation warning, and then launches the server. Because config.auth is absent, create_app() does not attach the API-key or JWT middleware. Unauthenticated requests can then reach the recipe API and, when enabled, /admin/reload. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56835 · Other software · MervinPraison · reported by GitHubHigh 8.3
PraisonAI is a multi-agent teams system. From version 3.11.0 to before version 4.6.59, PraisonAI's Slack bot applies its configured allowed_users, allowed_channels, and unknown-user pairing policy in the normal Slack message event handler, but not in the adjacent Slack app_mention event handler. A Slack workspace user who can mention the bot in a channel where the Slack app is present can trigger the configured PraisonAI agent even when: the sender is not in BotConfig.allowed_users; the channel is not in BotConfig.allowed_channels; unknown_user_policy="deny" is configured; and the same event content is correctly dropped by the normal message handler. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56834 · Other software · MervinPraison · reported by GitHubHigh 7.5
PraisonAI is a multi-agent teams system. From version 3.8.1 to before version 4.6.59, PraisonAI's Dynamic Context Discovery feature exposes artifact helper tools through ctx.get_tools(). The implemented artifact tools do not enforce that the supplied artifact_path is an artifact created by the configured store or that it lives under the configured artifact base directory. Instead, artifact_head, artifact_tail, artifact_grep, and artifact_chunk wrap the caller-supplied path directly into an ArtifactRef and then read it from the host filesystem. As a result, any prompt/user/tool-caller that can influence those tool arguments can read files readable by the PraisonAI process, such as project .env files, cloud credentials, SSH keys, source files, or other local data. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56833 · Other software · MervinPraison · reported by GitHubHigh 7.5
PraisonAI is a multi-agent teams system. From version 3.8.1 to before version 4.6.59, PraisonAI's Dynamic Context module provides filesystem-backed history and terminal-log storage. The module also exports agent-callable tool factories. Those tools accept run_id and agent_id arguments from the tool caller. The underlying stores join those values into filesystem paths without rejecting absolute paths or .. traversal. Because run_id can be an absolute path and agent_id can contain traversal, a lower-trust prompt/user that can call these tools can read .jsonl and .log files outside the configured Dynamic Context base directory. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026PraisonAI is a multi-agent teams system.CVE-2026-56832 · Other software · MervinPraison · reported by GitHubHigh 8.8
PraisonAI is a multi-agent teams system. From version 4.5.2 to before version 4.6.59, praisonai.bots.DiscordApproval approves a pending dangerous tool call when it sees any later non-bot message in the configured Discord channel whose text is classified as approval, such as yes. As a result, any user who can post in the configured approval channel can approve a pending high-risk tool call by sending yes after the approval message appears. the same PoV also shows that the Slack and Telegram messaging approval backends have no explicit approver allowlist parameter, but the primary report-grade issue is the Discord backend's unthreaded channel cross-talk: the approving message does not need to be a reply or otherwise request-bound. This issue has been patched in version 4.6.59.
Product
PraisonAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com
Oct 23, 2026AgenticSeek before commit 409840c contains a remote code execution vulnerability because llm_server/app.py hardcodes Flask debug mode while binding to all interCVE-2026-115727 · Other software · Fosowl · reported by VulnCheckCritical 9.2
AgenticSeek before commit 409840c contains a remote code execution vulnerability because llm_server/app.py hardcodes Flask debug mode while binding to all interfaces. Attackers who reach the server port and obtain the derivable Werkzeug debugger PIN can execute Python code and operating system commands with llm_server privileges.
Product
AgenticSeek
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026Chatwoot before 4.17.0 contains an authorization bypass vulnerability that allows authenticated agent bots to access other accounts by supplying a foreign contaCVE-2026-115725 · Other software · chatwoot · reported by VulnCheckLow 2.3
Chatwoot before 4.17.0 contains an authorization bypass vulnerability that allows authenticated agent bots to access other accounts by supplying a foreign contact inbox source_id. Attackers can call the conversations create endpoint with another account's source_id to create or retrieve conversations and post messages to that account's contacts.
Product
chatwoot
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 23, 2026Improper Certificate Validation vulnerability in the WebSocket streaming client in agentjido req_llm allows an on-path attacker to read and modify OpenAI WebSocCVE-2026-107685 · Other software · agentjido · reported by 6b3ad84c-e1a6-4bf7-a703-f496b71e49dbCritical 9.1
Improper Certificate Validation vulnerability in the WebSocket streaming client in agentjido req_llm allows an on-path attacker to read and modify OpenAI WebSocket sessions, including the authorization bearer token, via a TLS certificate the client does not trust. ReqLLM.Streaming.WebSocketSession.Client.start/3 in lib/req_llm/streaming/web_socket_session/client.ex starts WebSockex without TLS options, and WebSockex defaults to insecure: true, which connects with verify: :verify_none. The option allowlist also drops any caller-supplied cacerts or ssl_options, so an application cannot turn verification on. The OpenAI Responses, Realtime, and Codex WebSocket paths reach this client and send the API key or OAuth access token in the authorization header. HTTP streaming through Finch and the HTTP/2 duplex session through Mint verify certificates and are not affected. This issue affects req_llm: from 1.9.0 before 1.29.0.
Product
req_llm
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record cna.erlef.org github.com
Oct 23, 2026Missing Origin Validation in WebSockets (CWE-1385) and Permissive Cross-domain Policy with Untrusted Domains (CWE-942) in the development web server (ADK Web) iCVE-2026-107858 · Computers, phones & browsers · Google · reported by cve-coordination@google.comHigh 7.2
Missing Origin Validation in WebSockets (CWE-1385) and Permissive Cross-domain Policy with Untrusted Domains (CWE-942) in the development web server (ADK Web) in Google Agent Development Kit (ADK) for Java (com.google.adk:google-adk-dev versions 0.1.0 before 1.13.0 and com.google.adk:google-adk-maven-plugin versions 0.3.0 before 1.13.0) allow a remote attacker to control a locally running agent and read its data via a malicious web page visited by a developer while the server is running. The /run_live WebSocket endpoint accepts connections from any origin by default (hardcoded via setAllowedOrigins("*") in versions 0.1.0 through 1.5.0, and defaulting to * via adk.web.cors.origins in versions 1.6.0 through 1.12.0), and since version 0.4.0 the REST API returns Access-Control-Allow-Origin: * by default. Because the server has no authentication and requests originate from the developer's browser (so binding to localhost does not prevent exploitation), a cross-origin page can send messages to the agent (triggering registered tools), stream agent responses, and list, read, create, or delete sessions and artifacts. In versions 1.6.0 through 1.12.0, users can mitigate the issue by setting
Product
Agent Development Kit (ADK) for Java
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 22, 2026The A2A server component in Amazon Strands Agents SDK for Python before 1.43.0 and Amazon Strands Agents SDK for TypeScript before 1.5.0 does not properly isolaCVE-2026-115419 · Other software · AWS · reported by ff89ba41-3aa1-4d27-914a-91399e9639e5High 8.6
The A2A server component in Amazon Strands Agents SDK for Python before 1.43.0 and Amazon Strands Agents SDK for TypeScript before 1.5.0 does not properly isolate conversation state between callers. An authenticated remote user could read other tenants' conversation history and add persistent instructions that alter model behavior for subsequent callers via HTTP requests to the A2A endpoint using a different context_id. To mitigate this issue, users should upgrade to strands-agents 1.43.0 (Python) or @strands-agents/sdk 1.5.0 (TypeScript).
Product
strands-agents
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record aws.amazon.com github.com
Oct 22, 2026OpenBao before 2.6.3 contains a missing authorization check that allows attackers to stop Agent and Proxy services by omitting the X-Vault-Header.CVE-2026-115364 · Other software · openbao · reported by VulnCheckLow 2.1
OpenBao before 2.6.3 contains a missing authorization check that allows attackers to stop Agent and Proxy services by omitting the X-Vault-Header. When require_request_header is enabled, attackers can send requests via server-side request forgery to /agent/v1/quit or /proxy/v1/quit to shut down these services.
Product
openbao
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com vulncheck.com
Oct 22, 2026The Agent Zero Web UI before 1.15 does not correctly sanitize user-provided input in chat messages or file uploads.CVE-2026-97635 · Other software · Agent Zero · reported by 9119a7d8-5eab-497f-8521-727c672e3725High 8.6
The Agent Zero Web UI before 1.15 does not correctly sanitize user-provided input in chat messages or file uploads. When Agent Zero reproduces content in the chat interface, any injected JavaScript is executed in the context of the victim’s authenticated Agent Zero Web UI session.
Product
Agent Zero
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 22, 2026Missing authorization in the unit resources download route in Canonical Juju 4.0 through 4.0.15 allows a controller-authenticated principal (a unit agent, or a CVE-2026-93336 · Other software · Canonical · reported by security@ubuntu.comHigh 7.1
Missing authorization in the unit resources download route in Canonical Juju 4.0 through 4.0.15 allows a controller-authenticated principal (a unit agent, or a user with only login permission) to read any application's charm resource in any model, including container-registry credentials, via a crafted request naming an arbitrary unit.
Product
Juju
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 22, 2026A cleartext storage of sensitive information vulnerability in WatchGuard Endpoint Security's Full Disk Encryption feature for macOS allows a local, unprivilegedCVE-2026-96293 · Network & remote access · WatchGuard · reported by 5d1c2695-1a31-4499-88ae-e847036fd7e3Medium 6.9
A cleartext storage of sensitive information vulnerability in WatchGuard Endpoint Security's Full Disk Encryption feature for macOS allows a local, unprivileged user or process to obtain a managed computer's FileVault Personal Recovery Key by reading a persistently written diagnostic log file left readable by all local users.
Product
WatchGuard Agent
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record psirt.watchguard.com
Oct 22, 2026An incorrect permission assignment vulnerability in the WatchGuard Agent's log directory allows a local, authenticated non-administrative user to create, modifyCVE-2026-115253 · Network & remote access · WatchGuard · reported by 5d1c2695-1a31-4499-88ae-e847036fd7e3High 7.3
An incorrect permission assignment vulnerability in the WatchGuard Agent's log directory allows a local, authenticated non-administrative user to create, modify, append to, or delete the agent's log files and write other files into the directory. This allows an attacker to tamper with or destroy the agent's security logs and stage further local attacks by placing executable content in a directory owned by a privileged service.
Product
WatchGuard Agent
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record psirt.watchguard.com
Oct 21, 2026Robots::Validate versions from 0.4.3 before 0.4.10 for Perl report clients as imposter robots when User-Agent matcher state carries over from a previous requestCVE-2026-112921 · Other software · reported by 9b29abf9-4ab0-4765-b253-1875cd9b441e
Robots::Validate versions from 0.4.3 before 0.4.10 for Perl report clients as imposter robots when User-Agent matcher state carries over from a previous request. Each Robots::Validate object holds one pure-Perl Aho-Corasick search machine and feeds every User-Agent into it without resetting its state. A User-Agent that ends part-way through a robot name leaves the machine mid-match, so the next User-Agent checked by the same object can complete that name. That client's address is then checked against the robot's DNS rule, which it fails. A remote client can use this to make a site that blocks bad robots block other visitors where the prefix User-Agent string is a suffix of a robot being validated. Note that deployments with Algorithm::AhoCorasick::XS are also affected, since it was unused due to a different bug.
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com metacpan.org
Oct 21, 2026CrewAI crewAI Framework Agent Loading Unsafe Reflection Remote Code Execution Vulnerability.CVE-2026-92206 · Other software · CrewAI · reported by zdi-disclosures@trendmicro.comHigh 8.8
CrewAI crewAI Framework Agent Loading Unsafe Reflection Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of CrewAI crewAI. User interaction is required to exploit this vulnerability in that the target must load a malicious agent configuration from the repository. The specific flaw exists within the load_agent_from_repository function. The process does not properly restrict a user-supplied argument before using it to import a module. An attacker can leverage this vulnerability to execute code in the context of the service account. Was ZDI-CAN-28387.
Product
CrewAI
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record zerodayinitiative.com
Oct 21, 2026The webhook platform gateway in Nous Research Hermes Agent versions builds the session chat_id as webhook:{route_name}:{delivery_id} and re-derives the authorizCVE-2026-114463 · Other software · Nous Research · reported by cret@cert.org
The webhook platform gateway in Nous Research Hermes Agent versions builds the session chat_id as webhook:{route_name}:{delivery_id} and re-derives the authorizing route by splitting that string, so a route name containing a ":" resolves toolset authorization to a different, higher-privileged prefix route than the one authenticated, which allows a remote attacker with the target HMAC to gain that higher-privileged route's toolsets.
Product
Hermes Agent
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 21, 2026Appointment Booking: Does not verify that an authenticated agent-level user is assigned to a customer before returning or searching that customer's recordCVE-2026-105257 · WordPress plugins & themes · reported by WPScanLow 2.7
The Appointment Booking Plugin WordPress plugin before 5.7.5 does not verify that an authenticated agent-level user is assigned to a customer before returning or searching that customer's record, allowing any agent to read and enumerate the personal information of every customer on the site, including those managed by other agents.
Product
Appointment Booking
Fixed in
5.7.5
What to do
Update Appointment Booking to 5.7.5 or newer.
Full record and what to do →NVD record wpscan.com
Oct 20, 2026HKUDS Vibe-Trading from 0.1.0 before 0.1.7 contains a path traversal vulnerability that allows remote attackers to read sensitive server files through LLM agentCVE-2026-114106 · Other software · HKUDS · reported by VulnCheckHigh 8.7
HKUDS Vibe-Trading from 0.1.0 before 0.1.7 contains a path traversal vulnerability that allows remote attackers to read sensitive server files through LLM agent file-read tools. Attackers can prompt the agent to invoke read_document, which lacks sandbox checks, or safe_user_path-gated tools to leak /etc/shadow, /proc/self/environ API keys, and /root credentials.
Product
Vibe-Trading
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com vulncheck.com
Oct 20, 2026Trigger.dev 4.5.0 before 4.6.0 contains an information disclosure vulnerability in the Dashboard Agent input proxy that stores a delegated user-actor token in bCVE-2026-114070 · Other software · triggerdotdev · reported by VulnCheckMedium 5.1
Trigger.dev 4.5.0 before 4.6.0 contains an information disclosure vulnerability in the Dashboard Agent input proxy that stores a delegated user-actor token in browser-readable session input records. Attackers holding a chat's public session token can read the in records endpoint to recover the token from payload.metadata.userActorToken and replay it for environment JWT exchange.
Product
trigger.dev
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com vulncheck.com
Oct 20, 2026Vulnerability in Oracle Essbase (component: Java Based Agent).CVE-2026-91662 · Other software · Oracle Corporation · reported by secalert_us@oracle.comMedium 6.4
Vulnerability in Oracle Essbase (component: Java Based Agent). The supported version that is affected is 21.8.3.0.0. Difficult to exploit vulnerability allows low privileged attacker with logon to the infrastructure where Oracle Essbase executes to compromise Oracle Essbase. While the vulnerability is in Oracle Essbase, attacks may significantly impact additional products (scope change). Successful attacks of this vulnerability can result in unauthorized access to critical data or complete access to all Oracle Essbase accessible data as well as unauthorized update, insert or delete access to some of Oracle Essbase accessible data. CVSS 3.1 Base Score 6.4 (Confidentiality and Integrity impacts). CVSS Vector: (CVSS:3.1/AV:L/AC:H/PR:L/UI:N/S:C/C:H/I:L/A:N).
Product
Oracle Essbase
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in Oracle Essbase (component: Java Based Agent).CVE-2026-91650 · Other software · Oracle Corporation · reported by secalert_us@oracle.comLow 3.7
Vulnerability in Oracle Essbase (component: Java Based Agent). The supported version that is affected is 21.8.3.0.0. Difficult to exploit vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Essbase. Successful attacks of this vulnerability can result in unauthorized read access to a subset of Oracle Essbase accessible data. CVSS 3.1 Base Score 3.7 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N).
Product
Oracle Essbase
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in Oracle Essbase (component: Java Based Agent).CVE-2026-91649 · Other software · Oracle Corporation · reported by secalert_us@oracle.comMedium 6
Vulnerability in Oracle Essbase (component: Java Based Agent). The supported version that is affected is 21.8.3.0.0. Easily exploitable vulnerability allows high privileged attacker with logon to the infrastructure where Oracle Essbase executes to compromise Oracle Essbase. While the vulnerability is in Oracle Essbase, attacks may significantly impact additional products (scope change). Successful attacks of this vulnerability can result in unauthorized access to critical data or complete access to all Oracle Essbase accessible data. CVSS 3.1 Base Score 6.0 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:L/AC:L/PR:H/UI:N/S:C/C:H/I:N/A:N).
Product
Oracle Essbase
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in Oracle Essbase (component: Java Based Agent).CVE-2026-91648 · Other software · Oracle Corporation · reported by secalert_us@oracle.comMedium 5.3
Vulnerability in Oracle Essbase (component: Java Based Agent). The supported version that is affected is 21.8.3.0.0. Easily exploitable vulnerability allows unauthenticated attacker with network access via TCP to compromise Oracle Essbase. Successful attacks of this vulnerability can result in unauthorized read access to a subset of Oracle Essbase accessible data. CVSS 3.1 Base Score 5.3 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N).
Product
Oracle Essbase
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in Oracle Essbase (component: Java Based Agent).CVE-2026-91643 · Other software · Oracle Corporation · reported by secalert_us@oracle.comMedium 5.3
Vulnerability in Oracle Essbase (component: Java Based Agent). The supported version that is affected is 21.8.3.0.0. Easily exploitable vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Essbase. Successful attacks of this vulnerability can result in unauthorized read access to a subset of Oracle Essbase accessible data. CVSS 3.1 Base Score 5.3 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N).
Product
Oracle Essbase
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in the Oracle Middleware Common Libraries and Tools product of Oracle Fusion Middleware (component: Remote Diagnostic Agent).CVE-2026-91617 · Other software · Oracle Corporation · reported by secalert_us@oracle.comMedium 5.5
Vulnerability in the Oracle Middleware Common Libraries and Tools product of Oracle Fusion Middleware (component: Remote Diagnostic Agent). Supported versions that are affected are 12.2.1.4.0, 14.1.1.0.0, 14.1.2.0.0 and 15.1.1.0.0. Easily exploitable vulnerability allows low privileged attacker with logon to the infrastructure where Oracle Middleware Common Libraries and Tools executes to compromise Oracle Middleware Common Libraries and Tools. Successful attacks of this vulnerability can result in unauthorized access to critical data or complete access to all Oracle Middleware Common Libraries and Tools accessible data. CVSS 3.1 Base Score 5.5 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N).
Product
Oracle Middleware Common Libraries and Tools
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in the Oracle Middleware Common Libraries and Tools product of Oracle Fusion Middleware (component: Remote Diagnostic Agent).CVE-2026-91616 · Other software · Oracle Corporation · reported by secalert_us@oracle.comLow 3.7
Vulnerability in the Oracle Middleware Common Libraries and Tools product of Oracle Fusion Middleware (component: Remote Diagnostic Agent). Supported versions that are affected are 12.2.1.4.0 and 14.1.1.0.0. Difficult to exploit vulnerability allows unauthenticated attacker with network access via HTTP to compromise Oracle Middleware Common Libraries and Tools. Successful attacks of this vulnerability can result in unauthorized read access to a subset of Oracle Middleware Common Libraries and Tools accessible data. CVSS 3.1 Base Score 3.7 (Confidentiality impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:L/I:N/A:N).
Product
Oracle Middleware Common Libraries and Tools
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026Vulnerability in the Oracle Enterprise Manager Base Platform product of Oracle Enterprise Manager (component: Agent Next Gen).CVE-2026-101518 · Other software · Oracle Corporation · reported by secalert_us@oracle.comHigh 8.8
Vulnerability in the Oracle Enterprise Manager Base Platform product of Oracle Enterprise Manager (component: Agent Next Gen). Supported versions that are affected are 13.5 and 24.1. Easily exploitable vulnerability allows low privileged attacker with network access via HTTP to compromise Oracle Enterprise Manager Base Platform. Successful attacks of this vulnerability can result in takeover of Oracle Enterprise Manager Base Platform. CVSS 3.1 Base Score 8.8 (Confidentiality, Integrity and Availability impacts). CVSS Vector: (CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H).
Product
Oracle Enterprise Manager Base Platform
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record oracle.com
Oct 20, 2026A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol.CVE-2026-111765 · Other software · a2aproject · reported by GitHubHigh 7.5
A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol. Starting in version 1.3.0.Final and prior to version 1.3.2.Final, the AndroidA2AHttpClient synchronous POST path overrides the disabled-redirect setting in AndroidA2AHttpClient.createConnection() with the PostBuilder default followRedirects value of true, leaving a residual of the earlier credential-redirect fix. JSONRPCTransport and RestTransport add interceptor-produced authentication headers and call synchronous post() without disabling redirects. AuthInterceptor permits X-API-Key, API-Key, X-Auth-Token, and X-Authentication headers; Android's repackaged OkHttp strips Authorization on cross-authority redirects but retains these nonstandard headers, disclosing a complete configured API key or token to another authority. The additional exposure applies when redirect or front-door behavior is compromised separately from the credential-handling component, or those components operate in different trust domains; a completely malicious initial origin already receives the credential on the first request. Autonomous clients require no human action, while user-invoked workflows can require interaction. Explicit foll
Product
a2a-java
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 20, 2026A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol.CVE-2026-111764 · Other software · a2aproject · reported by GitHubHigh 7.5
A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol. Prior to 1.3.0.Final, the default InMemoryTaskStore configuration performs no authorization on task reads: the no-argument constructor sets authorizationProvider to null, and InMemoryTaskStore.get(String taskId) returns the stored task without checking the caller or owner. Authorization depends on supplying a TaskAuthorizationProvider and routing requests through AuthorizationRequestHandlerDecorator, whether installed through CDI or explicitly by the application. Default store wiring and manually constructed DefaultRequestHandler.builder().build() handlers, including integrations such as Spring, leave task reads unprotected when no effective authorization-provider and decorator check is installed. In a multi-caller deployment using the default configuration, a caller who knows or guesses another caller's task ID can retrieve the task status, context ID, message history, and artifacts. The default/reference transport endpoints do not enforce authentication, so the task-read disclosure is reachable by an unauthenticated caller rather than requiring a privileged account. This issue is fixed in version 1.3.0.Fina
Product
a2a-java
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 20, 2026A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol.CVE-2026-111763 · Other software · a2aproject · reported by GitHubMedium 5.7
A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol. Prior to 1.3.0.Final, a malicious or compromised remote agent can cause the default JDK-based client to forward a configured API key to a third-party server. When a client application configures credentials through `CredentialService`, `AuthInterceptor.intercept` in `client/transport/spi/src/main/java/org/a2aproject/sdk/client/transport/spi/interceptors/auth/AuthInterceptor.java` uses `APIKeySecurityScheme.name()` from the remote AgentCard directly as the HTTP header name containing the credential, without checking that the declared location is HEADER. JdkA2AHttpClient uses followRedirects(NORMAL), so an agent-controlled cross-origin redirect forwards the custom credential-bearing header even when the standard Authorization header is stripped. The attack requires configured API key credentials for the agent's security scheme and control of the agent's AgentCard and redirect response, but no additional client privileges. This issue is fixed in version 1.3.0.Final.
Product
a2a-java
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 20, 2026A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol.CVE-2026-111762 · Other software · a2aproject · reported by GitHubMedium 5.7
A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol. Prior to 1.3.0.Final, RestHandler.parseRequestBody in `transport/rest/src/main/java/org/a2aproject/sdk/transport/rest/handler/RestHandler.java` logs the entire JSON request body at SEVERE level when JsonFormat.parser().merge() throws InvalidProtocolBufferException. With push notifications enabled, createTaskPushNotificationConfiguration reaches this parser for requests to `/v1/tenants/{tenant}/tasks/{taskId}/pushNotificationConfigs`, whose token and authentication.credentials fields can contain bearer tokens and authentication credentials for client push endpoints. A valid JSON body that fails protobuf deserialization, including an unknown field under strict parsing or a schema mismatch, causes those secrets to be written to the server logs in cleartext. An attacker with access to the server logs or a log aggregation system can retrieve the credentials and potentially use the credentials to send forged task status updates to the client's push notification endpoint. This issue is fixed in version 1.3.0.Final.
Product
a2a-java
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 20, 2026A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol.CVE-2026-111761 · Other software · a2aproject · reported by GitHubMedium 5.8
A2A Java SDK is the official Java SDK for the Agent2Agent (A2A) Protocol. Prior to 1.3.0.Final, BasePushNotificationSender in server-common sends task data to the URL in TaskPushNotificationConfig without destination validation, scheme restrictions, or bounds on redirects. InMemoryPushNotificationConfigStore.setInfo stores the URL verbatim, and onCreateTaskPushNotificationConfig does not validate the destination. In the default deployment with push notifications enabled, an unauthenticated caller can register a destination through an inline SendMessage configuration or CreateTaskPushNotificationConfig for an existing task, causing task status and artifact updates to send requests from the server's network position to loopback, private, link-local cloud-metadata, IPv6 unique-local, unspecified, or multicast destinations where routing permits. JdkA2AHttpClient and VertxA2AHttpClient follow redirects, allowing a webhook redirect to bypass creation-time destination checks and convert a POST into a GET; hostname-only checks also do not prevent DNS rebinding. Direct POST delivery can push attacker-influenced task data into internal services, whereas a redirected POST-to-GET request drops
Product
a2a-java
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 19, 2026DB-GPT through 0.8.2 contains an unauthenticated remote code execution vulnerability because the /api/v1/chat/react-agent endpoint relies on a mock get_user_froCVE-2026-77932 · Other software · eosphoros-ai · reported by VulnCheckCritical 9.3
DB-GPT through 0.8.2 contains an unauthenticated remote code execution vulnerability because the /api/v1/chat/react-agent endpoint relies on a mock get_user_from_headers function that grants every caller the admin role. Unauthenticated attackers can send prompts that make the ReAct agent's unsandboxed code_interpreter and shell_interpreter tools run arbitrary Python or bash as root in the webserver container, including reaching the bundled database with default credentials.
Product
DB-GPT
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 19, 2026EvoScientist 0.1.0 before 0.2.5 contains an approval bypass vulnerability that allows attackers to run host shell commands by delegating them to async sub-agentCVE-2026-77930 · Other software · EvoScientist · reported by VulnCheckCritical 9.3
EvoScientist 0.1.0 before 0.2.5 contains an approval bypass vulnerability that allows attackers to run host shell commands by delegating them to async sub-agents lacking an interrupt_on gate. Attackers can plant instructions in ingested content such as fetched web pages, so the agent calls start_async_task and runs commands with user privileges and environment secrets.
Product
EvoScientist
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 19, 2026piflow 2.0.0 through 2.0.1 contains a remote code execution vulnerability in its DeepAgents chat agent, whose POST /chat and POST /chat/stream endpoints lack auCVE-2026-109442 · Other software · cas-bigdatalab · reported by VulnCheckCritical 9.3
piflow 2.0.0 through 2.0.1 contains a remote code execution vulnerability in its DeepAgents chat agent, whose POST /chat and POST /chat/stream endpoints lack authentication. Attackers can instruct the agent to call the exec_shell tool, which runs arbitrary commands via subprocess.run without human approval or an allow-list, exposing environment secrets such as LLM API keys and database credentials.
Product
piflow
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com github.com
Oct 19, 2026Cognition before 0.13.0 contains an unauthenticated remote code execution vulnerability due to defaulting its agent sandbox backend to local execution and shippCVE-2026-109437 · Other software · CognicellAI · reported by VulnCheckCritical 9.2
Cognition before 0.13.0 contains an unauthenticated remote code execution vulnerability due to defaulting its agent sandbox backend to local execution and shipping no authentication on its REST/SSE API. Attackers can create a session via POST /sessions and post a message via POST /sessions/{id}/messages without credentials, causing the Deep Agents ReAct loop to run attacker-directed shell commands through the local shell backend, resulting in remote code execution as the server process.
Product
Cognition
What to do
Apply the vendor’s update. The NVD record links to the advisory.
Full record and what to do →NVD record github.com lasso.security

Where this comes from: the exploited list is CISA’s Known Exploited Vulnerabilities catalog. Everything else is the whole National Vulnerability Database, where each record is published by the team that found the bug (Wordfence, Patchstack, WPScan, the vendors themselves and others); we show who. Exploit odds are FIRST’s EPSS, updated daily. Titles and categories are ours, so check the linked record before acting on a detail. For a write-up of the big ones, see our advisories. Every vendor, and every WordPress plugin and theme with a recorded flaw, also has its own page. All of it is free to reuse as JSON, CSV or RSS.

Weekly security brief

The flaws that hit local websites, every Friday.

One short email: what got exploited this week, which WordPress plugins broke, and what to check. Unsubscribe in one click.

Vulnerability database

The flaws that hit local websites, every Friday.

Tell us what you're building.