Claude Code is an agentic coding tool. Prior to version 2.1.64, Claude Code's sandbox did not prevent sandboxed processes from creating symlinks pointing to locations outside the workspace. When Claude Code subsequently wrote to a path within such a symlink, its unsandboxed process followed the symlink and wrote to the target location outside the workspace without prompting the user for confirmation. This allowed a sandbox escape where neither the sandboxed command nor the unsandboxed app could independently write outside the workspace, but their combination could write to arbitrary locations, potentially leading to code execution outside the sandbox. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window to trigger sandboxed code execution via prompt injection. Users on standard Claude Code auto-update have received this fix automatically. Users performing manual updates are advised to update to version 2.1.64 or later.

Claude Code is an agentic coding tool. Prior to version 2.0.72, due to an error in command parsing, it was possible to bypass the Claude Code confirmation prompt to trigger execution of untrusted commands through the find command. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window. This issue has been patched in version 2.0.72.

Claude Code is an agentic coding tool. Prior to version 2.0.55, Claude Code failed to properly validate commands using piped sed operations with the echo command, allowing attackers to bypass file write restrictions. This vulnerability enabled writing to sensitive directories like the .claude folder and paths outside the project scope. Exploiting this required the ability to execute commands through Claude Code with the "accept edits" feature enabled. This issue has been patched in version 2.0.55.

Claude Code is an agentic coding tool. Prior to version 2.0.74, due to a Bash command validation flaw in parsing ZSH clobber syntax, it was possible to bypass directory restrictions and write files outside the current working directory without user permission prompts. Exploiting this required the user to use ZSH and the ability to add untrusted content into a Claude Code context window. This issue has been patched in version 2.0.74.

Claude Code is an agentic coding tool. Prior to version 2.0.57, Claude Code failed to properly validate directory changes when combined with write operations to protected folders. By using the cd command to navigate into sensitive directories like .claude, it was possible to bypass write protection and create or modify files without user confirmation. Reliably exploiting this required the ability to add untrusted content into a Claude Code context window. This issue has been patched in version 2.0.57.

Claude Code is an agentic coding tool. In versions below 0.2.111, a path validation flaw using prefix matching instead of canonical path comparison, makes it possible to bypass directory restrictions and access files outside the CWD. Successful exploitation depends on the presence of (or ability to create) a directory with the same prefix as the CWD and the ability to add untrusted content into a Claude Code context window. This is fixed in version 0.2.111.

In versions 2.1.63 through 2.1.83 of Claude Code, the folder trust determination logic used the git worktree commondir file without validating its contents. An attacker could craft a malicious repository with a commondir file pointing to a path the victim had previously trusted, causing Claude Code to bypass its trust confirmation dialog and immediately execute hooks defined in `.claude/settings.json`. Exploitation requires the victim to clone the malicious repository and run Claude Code within it, and the attacker must know or guess a path the victim had already trusted. This issue has been fixed in version 2.1.84.

Claude Code is an agentic coding tool. Versions prior to 2.1.53 resolved the permission mode from settings files, including the repo-controlled .claude/settings.json, before determining whether to display the workspace trust confirmation dialog. A malicious repository could set permissions.defaultMode to bypassPermissions in its committed .claude/settings.json, causing the trust dialog to be silently skipped on first open. This allowed a user to be placed into a permissive mode without seeing the trust confirmation prompt, making it easier for an attacker-controlled repository to gain tool execution without explicit user consent. This issue has been patched in version 2.1.53.

Claude Code is an agentic coding tool. Prior to version 2.1.2, Claude Code's bubblewrap sandboxing mechanism failed to properly protect the .claude/settings.json configuration file when it did not exist at startup. While the parent directory was mounted as writable and .claude/settings.local.json was explicitly protected with read-only constraints, settings.json was not protected if it was missing. This allowed malicious code running inside the sandbox to create this file and inject persistent hooks (such as SessionStart commands) that would execute with host privileges when Claude Code was restarted. This issue has been patched in version 2.1.2.

Claude Code is an agentic coding tool. Prior to Claude Code version 1.0.39, when using Claude Code with Yarn versions 2.0+, Yarn plugins are auto-executed when running yarn --version. This could lead to a bypass of the directory trust dialog in Claude Code, as plugins would be executed prior to the user accepting the risks of working in an untrusted directory. Users running Yarn Classic were unaffected by this issue. This issue has been fixed in version 1.0.39. Users on standard Claude Code auto-update will have received this fix automatically. Users performing manual updates are advised to update to the latest version.

Claude Code is an agentic coding tool. Prior to version 1.0.39, when running on a machine with Yarn 3.0 or above, Claude Code could have been tricked to execute code contained in a project via yarn plugins before the user accepted the startup trust dialog. Exploiting this would have required a user to start Claude Code in an untrusted directory and to be using Yarn 3.0 or above. This issue has been patched in version 1.0.39.

Claude Code is an agentic coding tool. Versions before 1.0.111 were vulnerable to Code Injection due to a bug in the startup trust dialog implementation. Claude Code could be tricked to execute code contained in a project before the user accepted the startup trust dialog. Exploiting this requires a user to start Claude Code in an untrusted directory. Users on standard Claude Code auto-update will have received this fix automatically. Users performing manual updates are advised to update to the latest version. This issue is fixed in version 1.0.111.

Claude Code is an agentic coding tool. From 2.1.59 until 2.1.128, the Claude Code /copy command wrote responses to a hardcoded, predictable path (/tmp/claude/response.md) without UID isolation, randomness, or symlink protection. The file was created world-readable (0644) in a world-traversable directory (0755), allowing any local user to read a privileged user's Claude response, which could contain secrets or credentials. Additionally, because the path was static and predictable, a local attacker could pre-create the directory and plant a symlink at the expected file path, causing the privileged process to follow the symlink and overwrite an attacker-chosen file with the response text. Exploiting this required a local unprivileged user on the same system and a privileged user to run the /copy command. This vulnerability is fixed in 2.1.128.

The Claude Desktop app gives you Claude Code with a graphical interface built for running multiple sessions side by side. Prior to 1.3834.0, the CoworkVMService component in Claude Desktop for Windows ran as SYSTEM and did not validate whether the VM bundle directory was a real directory or an NTFS directory junction before creating files within it. A local non-elevated user could replace the user-writable VM bundle directory with a directory junction pointing to an attacker-chosen location, causing the service to create a SYSTEM-owned file in an arbitrary directory. This could be leveraged for local privilege escalation. This vulnerability is fixed in 1.3834.0.

The Claude SDK for Python provides access to the Claude API from Python applications. From version 0.86.0 to before version 0.87.0, the async local filesystem memory tool in the Anthropic Python SDK validated that model-supplied paths resolved inside the sandboxed memory directory, but then returned the unresolved path for subsequent file operations. A local attacker able to write to the memory directory could retarget a symlink between validation and use, causing reads or writes to escape the sandbox. The synchronous memory tool implementation was not affected. This issue has been patched in version 0.87.0.

Claude SDK for TypeScript provides access to the Claude API from server-side TypeScript or JavaScript applications. From version 0.79.0 to before version 0.81.0, the local filesystem memory tool in the Anthropic TypeScript SDK validated model-supplied paths using a string prefix check that did not append a trailing path separator. A model steered by prompt injection could supply a crafted path that resolved to a sibling directory sharing the memory root's name as a prefix, allowing reads and writes outside the sandboxed memory directory. This issue has been patched in version 0.81.0.

Claude Code is an agentic coding tool. Prior to version 2.0.31, due to an error in sed command parsing, it was possible to bypass the Claude Code read-only validation and write to arbitrary files on the host system. This issue has been patched in version 2.0.31.

Claude Code is an agentic coding tool. Prior to version 1.0.4, it's possible to bypass the Claude Code confirmation prompts to read a file and then send file contents over the network without user confirmation due to an overly broad allowlist of safe commands. Reliably exploiting this requires the ability to add untrusted content into a Claude Code context window. Users on standard Claude Code auto-update received this fix automatically after release. Current users of Claude Code are unaffected, as versions prior to 1.0.24 are deprecated and have been forced to update.

Claude Code is an agentic coding tool. In versions below 1.0.20, an error in command parsing makes it possible to bypass the Claude Code confirmation prompt to trigger execution of an untrusted command. Reliably exploiting this requires the ability to add untrusted content into a Claude Code context window. This is fixed in version 1.0.20.

Markdown Preview Enhanced before 0.8.28 opens external files and links from the preview through a shell and does not validate untrusted inputs taken from the markdown document - the diagram filename attribute, imported file paths, and the latex_engine code-chunk attribute. On Windows, a crafted markdown document can inject operating system commands that execute when the document is previewed. Fixed in 0.8.28 by passing these inputs as literal arguments instead of through a shell and validating them before use.

In Zmanda Management Console 3.3.9, ZMC_Admin_Advanced?form=adminTasks&action=Apply&command= allows CSRF, as demonstrated by command injection with shell metacharacters. This may depend on weak default credentials.

Pi-Hole 4.3 allows Command Injection.

Warp is an agentic development environment. From 0.2023.03.21.08.02.stable_00 until 0.2026.05.06.15.42.stable_01, Warp contains a command injection issue in the legacy SSH background command path. Warp used the remote working directory reported by the session when building helper commands for SSH-backed metadata collection. A remote host, repository, or directory name controlled by an attacker could cause that helper command to execute additional shell syntax on the remote host as the victim's authenticated SSH account. This vulnerability is fixed in 0.2026.05.06.15.42.stable_01.

Inductive Automation Ignition getJavaExecutable Directory Traversal Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Inductive Automation Ignition. User interaction is required to exploit this vulnerability in that the target must connect to a malicious server. The specific flaw exists within the getJavaExecutable method. The issue results from the lack of proper validation of a user-supplied path prior to using it in file operations. An attacker can leverage this vulnerability to execute code in the context of the current user. Was ZDI-CAN-22029.

Improper neutralization of special elements used in an OS command ('OS command injection') vulnerability in TUBITAK BILGEM Software Technologies Research Institute Pardus OS My Computer allows OS Command Injection. This issue affects Pardus OS My Computer: from <=0.7.5 before 0.8.0.

File::Find::Rule through 0.34 for Perl is vulnerable to Arbitrary Code Execution when `grep()` encounters a crafted filename. A file handle is opened with the 2 argument form of `open()` allowing an attacker controlled filename to provide the MODE parameter to `open()`, turning the filename into a command to be executed. Example: $ mkdir /tmp/poc; echo > "/tmp/poc/|id" $ perl -MFile::Find::Rule \     -E 'File::Find::Rule->grep("foo")->in("/tmp/poc")' uid=1000(user) gid=1000(user) groups=1000(user),100(users)

Tabby (formerly Terminus) is a highly configurable terminal emulator. Prior to 1.0.233, Tabby registers itself as the handler for the tabby:// URL scheme on all platforms. The URL scheme handler supports a run command that directly executes OS commands with no user confirmation, sanitization, or sandboxing. An attacker can craft a malicious link (tabby://run?command=...) and deliver it via a website, email, chat message, or any other medium. When a victim clicks the link, the OS launches Tabby which immediately spawns the specified command as a child process with the user's full privileges. This is a zero-click-after-link-visit RCE vulnerability. This vulnerability is fixed in 1.0.233.

radare2-mcp version 1.6.0 and earlier contains an os command injection vulnerability that allows remote attackers to execute arbitrary commands by bypassing the command filter through shell metacharacters in user-controlled input passed to r2_cmd_str(). Attackers can inject shell metacharacters through the jsonrpc interface parameters to achieve remote code execution on the host running radare2-mcp without requiring authentication.

An issue was discovered in Centreon before 18.10.8, 19.10.1, and 19.04.2. It allows CSRF with resultant remote command execution via shell metacharacters in a POST to centreon-autodiscovery-server/views/scan/ajax/call.php in the Autodiscovery plugin.

Dulwich is a pure-Python implementation of the Git file formats and protocols. Versions starting with 0.10.0 and prior to 1.2.5 have an arbitrary file write leading to remote code execution when cloning or checking out a malicious Git repository on Windows. Dulwich's path-element validator accepted tree entries whose filenames contained bytes that Windows interprets as structural path syntax. Contributing configuration bugs made matters worse. The core.protectNTFS and core.protectHFS settings were looked up under a wrong option name and so user-set values were silently ignored, and core.protectNTFS only defaulted to true on Windows (Git upstream has defaulted it to true everywhere since CVE-2019-1353). Both have been corrected. Anyone who clones, fetches, or checks out an untrusted repository with Dulwich on Windows - either through the Dulwich CLI, porcelain.clone, or any downstream tool built on Dulwich - is impacted. POSIX clones are not directly exploitable (on POSIX \ is a literal filename byte), but a POSIX user can unknowingly propagate a malicious tree to Windows consumers via push or re-publication. This issue is fixed in Dulwich 1.2.5. Users should upgrade to 1.2.5 or later. There is no effective pre-patch workaround. On affected versions the core.protectNTFS configuration key was silently ignored, so setting it to true does not mitigate the issue. Users who cannot upgrade should avoid cloning, fetching, or checking out untrusted repositories with Dulwich on Windows. After upgrading the NTFS validator is on by default on every platform, so no additional configuration is required.

Directory Traversal vulnerability in the extractFile method of org.codehaus.plexus.util.Expand in plexus-utils before 6d780b3378829318ba5c2d29547e0012d5b29642. This allows an attacker to execute arbitrary code

BentoML is a Python library for building online serving systems optimized for AI apps and model inference. Prior to 1.4.39, src/bentoml/_internal/container/frontend/dockerfile/templates/base_v2.j2 interpolates docker.base_image raw with no escaping, newline filtering, or validation. A malicious bento.yaml with a multi-line docker.base_image value smuggles arbitrary Dockerfile directives into the generated Dockerfile, and bentoml containerize then runs docker build which executes the injected RUN directives on the victim host. This vulnerability is fixed in 1.4.39.

@rtk-ai/rtk-rewrite transparently rewrites shell commands executed via OpenClaw's exec tool to their RTK equivalents. In 1.0.0, the @rtk-ai/rtk-rewrite OpenClaw plugin passes attacker-controlled input directly into a shell-backed execSync() template string without shell-safe escaping. JSON.stringify() wraps the value in double quotes and escapes inner double-quotes and backslashes, but leaves $() and backtick shell metacharacters untouched. Because execSync delegates execution to /bin/sh -c, the shell expands $(...) substitutions even inside double-quoted strings, causing the injected subcommand to execute before rtk is invoked. An attacker who can influence the exec tool's command parameter (e.g., via an LLM agent prompt or gateway/tool-call input) achieves arbitrary OS command execution with the privileges of the plugin/gateway process.

BentoML is a Python library for building online serving systems optimized for AI apps and model inference. Prior to 1.4.39, a malicious bentofile.yaml containing a newline-injected value in envs[*].name produces unquoted RUN directives in the BentoML-generated Dockerfile. When the victim runs bentoml containerize on the imported bento, those RUN directives execute on the host during docker build. This vulnerability is fixed in 1.4.39.

Ghidra versions prior to 12.0.3 improperly process annotation directives embedded in automatically extracted binary data, resulting in arbitrary command execution when an analyst interacts with the UI. Specifically, the @execute annotation (which is intended for trusted, user-authored comments) is also parsed in comments generated during auto-analysis (such as CFStrings in Mach-O binaries). This allows a crafted binary to present seemingly benign clickable text which, when clicked, executes attacker-controlled commands on the analyst’s machine.

Improper link resolution before file access ('link following') vulnerability in TUBITAK BILGEM Software Technologies Research Institute Pardus About allows Symlink Attack. This issue affects Pardus About: before 1.2.2.

A path traversal vulnerability was identified in GitHub Enterprise Server management console that allowed the bypass of CSRF protections. This could potentially lead to privilege escalation. To exploit this vulnerability, an attacker would need to target a user that was actively logged into the management console. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.5 and was fixed in versions 3.1.19, 3.2.11, 3.3.6, 3.4.1. This vulnerability was reported via the GitHub Bug Bounty program.

NI FlexLogger usiReg URI File Parsing Directory Traversal Remote Code Execution Vulnerability. This vulnerability allows remote attackers to create arbitrary files on affected installations of NI FlexLogger. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file. The specific flaw exists within the parsing of URI files by the usiReg component. The issue results from the lack of proper validation of a user-supplied path prior to using it in file operations. An attacker can leverage this vulnerability to execute code in the context of the current user. Was ZDI-CAN-21805.

A potential vulnerability was reported in Lenovo PCManager prior to version 5.0.10.4191 that may allow code execution when visiting a specially crafted website.

Dulwich is a pure-Python implementation of the Git file formats and protocols. Starting in version 0.24.0 and prior to version 1.2.5, Dulwich's `ProcessMergeDriver` substitutes the file path (from the git tree, controllable by an attacker via a malicious branch) into the merge driver command via the `%P` placeholder and executes it with `subprocess.run(..., shell=True)`. An attacker who can cause a victim to merge an untrusted branch can achieve arbitrary command execution by crafting malicious file paths. Version 1.2.5 fixes the issue.

GNU Bash through 4.3 bash43-026 does not properly parse function definitions in the values of environment variables, which allows remote attackers to execute arbitrary commands via a crafted environment, as demonstrated by vectors involving the ForceCommand feature in OpenSSH sshd, the mod_cgi and mod_cgid modules in the Apache HTTP Server, scripts executed by unspecified DHCP clients, and other situations in which setting the environment occurs across a privilege boundary from Bash execution. NOTE: this vulnerability exists because of an incomplete fix for CVE-2014-6271, CVE-2014-7169, and CVE-2014-6277.

An OS command injection vulnerability in the web management interface of certain ASUS router models allows remote authenticated administrators to execute arbitrary system commands via a crafted parameter. Refer to the 'Security Update for ASUS Router Firmware' section on the ASUS Security Advisory for more information.

Path Traversal in Clasp impacting versions < 3.2.0 allows a remote attacker to perform remote code execution via a malicious Google Apps Script project containing specially crafted filenames with directory traversal sequences.

The Drag and Drop Multiple File Upload for Contact Form 7 plugin for WordPress is vulnerable to arbitrary file deletion due to insufficient file path validation in the 'dnd_remove_uploaded_files' function in all versions up to, and including, 1.3.8.7. This makes it possible for unauthenticated attackers to add arbitrary file paths (such as ../../../../wp-config.php) to uploaded files on the server, which can easily lead to remote code execution when an Administrator deletes the message. Exploiting this vulnerability requires the Flamingo plugin to be installed and activated.

Session fixation in Visual Studio Code allows an unauthorized attacker to elevate privileges over a network.

Composer is a dependency manager for PHP. Versions 1.0 through 2.2.26 and 2.3 through 2.9.5 contain a command injection vulnerability in the Perforce::syncCodeBase() method, which appends the $sourceReference parameter to a shell command without proper escaping, and additionally in the Perforce::generateP4Command() method as in GHSA-wg36-wvj6-r67p / CVE-2026-40176, which interpolates user-supplied Perforce connection parameters (port, user, client) from the source url field without proper escaping. An attacker can inject arbitrary commands through crafted source reference or source url values containing shell metacharacters, even if Perforce is not installed. Unlike CVE-2026-40176, the source reference and url are provided as part of package metadata, meaning any compromised or malicious Composer repository can serve package metadata declaring perforce as a source type with malicious values. This vulnerability is exploitable when installing or updating dependencies from source, including the default behavior when installing dev-prefixed versions. This issue has been fixed in Composer 2.2.27 (2.2 LTS) and 2.9.6 (mainline). If developers are unable to immediately update, they can avoid installing dependencies from source by using --prefer-dist or the preferred-install: dist config setting, and only use trusted Composer repositories as a workaround.

Let's Encrypt client and ACME library written in Go (Lego). Prior to 4.34.0, the webroot HTTP-01 challenge provider in lego is vulnerable to arbitrary file write and deletion via path traversal. A malicious ACME server can supply a crafted challenge token containing ../ sequences, causing lego to write attacker-influenced content to any path writable by the lego process. This vulnerability is fixed in 4.34.0.

The Career Section plugin for WordPress is vulnerable to Cross-Site Request Forgery leading to Path Traversal and Arbitrary File Deletion in all versions up to, and including, 1.6. This is due to missing nonce validation and insufficient file path validation on the delete action in the 'appform_options_page_html' function. This makes it possible for unauthenticated attackers to delete arbitrary files on the server via a forged request, granted they can trick a site administrator into performing an action such as clicking on a link.

PraisonAIAgents is a multi-agent teams system. Prior to 1.5.128, he memory hooks executor in praisonaiagents passes a user-controlled command string directly to subprocess.run() with shell=True at src/praisonai-agents/praisonaiagents/memory/hooks.py. No sanitization is performed and shell metacharacters are interpreted by /bin/sh before the intended command executes. Two independent attack surfaces exist. The first is via pre_run_command and post_run_command hook event types registered through the hooks configuration. The second and more severe surface is the .praisonai/hooks.json lifecycle configuration, where hooks registered for events such as BEFORE_TOOL and AFTER_TOOL fire automatically during agent operation. An agent that gains file-write access through prompt injection can overwrite .praisonai/hooks.json and have its payload execute silently at every subsequent lifecycle event without further user interaction. This vulnerability is fixed in 1.5.128.

PraisonAI is a multi-agent teams system. Prior to 4.5.128, cmd_unpack in the recipe CLI extracts .praison tar archives using raw tar.extract() without validating archive member paths. A .praison bundle containing ../../ entries will write files outside the intended output directory. An attacker who distributes a malicious bundle can overwrite arbitrary files on the victim's filesystem when they run praisonai recipe unpack. This vulnerability is fixed in 4.5.128.