PraisonAI is a multi-agent teams system. Prior to praisonai version 4.6.9 and praisonaiagents version 1.6.9, the fix for CVE-2026-40315 added input validation to SQLiteConversationStore only. Nine sibling backends — MySQL, PostgreSQL, async SQLite/MySQL/PostgreSQL, Turso, SingleStore, Supabase, SurrealDB — pass table_prefix straight into f-string SQL. Same root cause, same code pattern, same exploitation. 52 unvalidated injection points across the codebase. postgres.py additionally accepts an unvalidated schema parameter used directly in DDL. This issue has been patched in praisonai version 4.6.9 and praisonaiagents version 1.6.9.

PraisonAI is a multi-agent teams system. Prior to 4.5.128, deploy.py constructs a single comma-delimited string for the gcloud run deploy --set-env-vars argument by directly interpolating openai_model, openai_key, and openai_base without validating that these values do not contain commas. gcloud uses a comma as the key-value pair separator for --set-env-vars. A comma in any of the three values causes gcloud to parse the trailing text as additional KEY=VALUE definitions, injecting arbitrary environment variables into the deployed Cloud Run service. This vulnerability is fixed in 4.5.128.

PraisonAI before 4.6.78 contains a path traversal vulnerability in ContextGatherer that fails to validate include paths in .praisoncontext and .praisoninclude files. Attackers can supply absolute paths or parent directory traversal sequences to read arbitrary files outside the workspace and include their contents in the generated context bundle.

PraisonAI before 4.6.78 contains arbitrary file write and command execution vulnerabilities in the AICoder component due to missing path validation and command sanitization in LLM tool calls. Attackers can inject malicious prompts through the chat interface to write files to arbitrary filesystem locations and execute arbitrary shell commands with root privileges.

PraisonAI (pip package praisonaiagents) before 1.6.78 automatically loads defaults from a project-local .praisonai/config.toml when constructing an Agent, and does not validate the defaults.output.output_file path. A repository-controlled config file can set output_file to an absolute or '..' traversal path; when the developer subsequently calls agent.start() without explicitly passing an output parameter, PraisonAI writes the agent response to that path (creating parent directories as needed), allowing an untrusted checked-out project to overwrite files outside the project root with the privileges of the user running PraisonAI.

PraisonAI (praisonaiagents) before 1.6.78 contains a path traversal vulnerability in the FastContext feature (praisonaiagents.context.fast). FastContextAgent.execute_tool() prepends the configured workspace_path only for relative paths and neither rejects absolute paths nor canonicalizes joined paths before enforcing workspace containment. As a result, tool arguments or model-generated function calls to grep_search, glob_search, read_file, or list_directory can supply absolute paths or '../' traversal sequences to read, search, and enumerate files outside the intended workspace directory, with file contents returned to the caller or injected into the model's tool-result context.

PraisonAI is a multi-agent teams system. Prior to version 4.6.34, PraisonAI's MCP (Model Context Protocol) server (praisonai mcp serve) registers four file-handling tools by default — praisonai.rules.create, praisonai.rules.show, praisonai.rules.delete, and praisonai.workflow.show. Each accepts a path or filename string from MCP tools/call arguments and joins it onto ~/.praison/rules/ (or, for workflow.show, accepts an absolute path) with no containment check. The JSON-RPC dispatcher passes params["arguments"] blind to each handler via **kwargs without validating against the advertised input schema. By setting rule_name="../../<some-path>" an attacker walks out of the rules directory and writes any file the running user can write. Dropping a Python .pth file into the user site-packages directory escalates this primitive to arbitrary code execution in any subsequent Python process the user spawns — the next praisonai CLI invocation, an IDE script run, the user's python REPL, or any background Python service. This issue has been patched in version 4.6.34.

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.

PraisonAI is a multi-agent teams system. Prior to 1.5.113, the Action Orchestrator feature contains a Path Traversal vulnerability that allows an attacker (or compromised agent) to write to arbitrary files outside of the configured workspace directory. By supplying relative path segments (../) in the target path, malicious actions can overwrite sensitive system files or drop executable payloads on the host. This vulnerability is fixed in 1.5.113.

PraisonAI is a multi-agent teams system. Prior to 1.5.113, PraisonAI's recipe registry pull flow extracts attacker-controlled .praison tar archives with tar.extractall() and does not validate archive member paths before extraction. A malicious publisher can upload a recipe bundle that contains ../ traversal entries and any user who later pulls that recipe will write files outside the output directory they selected. This is a path traversal / arbitrary file write vulnerability on the client side of the recipe registry workflow. It affects both the local registry pull path and the HTTP registry pull path. The checksum verification does not prevent exploitation because the malicious traversal payload is part of the signed bundle itself. This vulnerability is fixed in 1.5.113.

PraisonAI is a multi-agent teams system. Prior to 1.5.113, _validate_path() calls os.path.normpath() first, which collapses .. sequences, then checks for '..' in normalized. Since .. is already collapsed, the check always passes. This makes the check completely useless and allows trivial path traversal to any file on the system. This vulnerability is fixed in 1.5.113.

PraisonAI before 4.6.78 fails to validate file path references in custom command templates, allowing attackers to read files outside the workspace. Attackers can include path traversal sequences like @../outside_secret.txt or absolute paths in project command files to exfiltrate process-readable files into model prompts.

PraisonAI is a multi-agent teams system. Prior to version 4.6.37, the _safe_extractall helper that all recipe pull, recipe publish, and recipe unpack flows route through validates each archive member's name for absolute paths, .. segments, and resolved-path escape — but does not validate member.linkname, does not reject symlink/hardlink members, and calls tar.extractall(dest_dir) without filter="data". A bundle that contains a symlink with a name inside dest_dir but a linkname pointing outside it, followed by a regular file whose path traverses through the just-created symlink, escapes dest_dir and lets the attacker write arbitrary content to an attacker-chosen location on the victim's filesystem. This issue has been patched in version 4.6.37.

PraisonAIAgents is a multi-agent teams system. Prior to 1.5.128, he list_files() tool in FileTools validates the directory parameter against workspace boundaries via _validate_path(), but passes the pattern parameter directly to Path.glob() without any validation. Since Python's Path.glob() supports .. path segments, an attacker can use relative path traversal in the glob pattern to enumerate arbitrary files outside the workspace, obtaining file metadata (existence, name, size, timestamps) for any path on the filesystem. This vulnerability is fixed in 1.5.128.

PraisonAI is a multi-agent teams system. Prior to 1.5.113, The PraisonAI templates installation feature is vulnerable to a "Zip Slip" Arbitrary File Write attack. When downloading and extracting template archives from external sources (e.g., GitHub), the application uses Python's zipfile.extractall() without verifying if the files within the archive resolve outside of the intended extraction directory. This vulnerability is fixed in 1.5.113.

PraisonAI is a multi-agent teams system. Prior to 1.5.113, PraisonAI's recipe registry publish endpoint writes uploaded recipe bundles to a filesystem path derived from the bundle's internal manifest.json before it verifies that the manifest name and version match the HTTP route. A malicious publisher can place ../ traversal sequences in the bundle manifest and cause the registry server to create files outside the configured registry root even though the request is ultimately rejected with HTTP 400. This is an arbitrary file write / path traversal issue on the registry host. It affects deployments that expose the recipe registry publish flow. If the registry is intentionally run without a token, any network client that can reach the service can trigger it. If a token is configured, any user with publish access can still exploit it. This vulnerability is fixed in 1.5.113.

Billy is an interface filesystem abstraction for Go. Prior to 5.9.0, multiple path traversal issues exist across different components of go-billy. Insufficient path sanitization and boundary enforcement may allow crafted paths (e.g., using ..) to escape intended base directories. While go-billy was not originally designed to provide a strong security boundary, some of these issues were inconsistent across some of the built-in implementations. This results in scenarios where applications relying on go-billy for some level of isolation may inadvertently expose access to unintended filesystem locations. This vulnerability is fixed in 5.9.0.

A vulnerability was identified in Consul and Consul Enterprise (“Consul”) such that using URL paths in L7 traffic intentions could bypass HTTP request path-based access rules.

The Kubernetes kubelet component allows arbitrary command execution via specially crafted gitRepo volumes.This issue affects kubelet: through 1.28.11, from 1.29.0 through 1.29.6, from 1.30.0 through 1.30.2.

The BuddyPress plugin for WordPress is vulnerable to Directory Traversal in all versions up to, and including, 14.1.0 via the id parameter. This makes it possible for authenticated attackers, with Subscriber-level access and above, to perform actions on files outside of the originally intended directory and enables file uploads to directories outside of the web root. Depending on server configuration it may be possible to upload files with double extensions.

LLaMA-Factory is a tuning library for large language models. Prior to version 0.9.4, a Server-Side Request Forgery (SSRF) vulnerability in the chat API allows any authenticated user to force the server to make arbitrary HTTP requests to internal and external networks. This can lead to the exposure of sensitive internal services, reconnaissance of the internal network, or interaction with third-party services. The same mechanism also allows for a Local File Inclusion (LFI) vulnerability, enabling users to read arbitrary files from the server's filesystem. The vulnerability exists in the `_process_request` function within `src/llamafactory/api/chat.py.` This function is responsible for processing incoming multimodal content, including images, videos, and audio provided via URLs. The function checks if the provided URL is a base64 data URI or a local file path (`os.path.isfile`). If neither is true, it falls back to treating the URL as a web URI and makes a direct HTTP GET request using `requests.get(url, stream=True).raw` without any validation or sanitization of the URL. Version 0.9.4 fixes the underlying issue.

A vulnerability in the web-based management interface of Cisco Firepower Management Center (FMC) Software could allow an authenticated, remote attacker to perform a directory traversal attack on an affected device. The attacker would require valid device credentials. The vulnerability is due to insufficient input validation of the HTTPS URL by the web-based management interface. An attacker could exploit this vulnerability by sending a crafted HTTPS request that contains directory traversal character sequences to an affected device. A successful exploit could allow the attacker to read or write arbitrary files on the device.

IBM SOAR QRadar Plugin App 1.0 through 5.0.3 could allow a remote attacker to traverse directories on the system. An attacker could send a specially crafted URL request containing "dot dot" sequences (/../) to view arbitrary files on the system. IBM X-Force ID: 260575.

An issue in the size query parameter (/views/file.py) of Austrian Archaeological Institute Openatlas before v8.12.0 allows attackers to execute a path traversal via a crafted request.

In Open-iSCSI tcmu-runner 1.3.x, 1.4.x, and 1.5.x through 1.5.2, xcopy_locate_udev in tcmur_cmd_handler.c lacks a check for transport-layer restrictions, allowing remote attackers to read or write files via directory traversal in an XCOPY request. For example, an attack can occur over a network if the attacker has access to one iSCSI LUN. NOTE: relative to CVE-2020-28374, this is a similar mistake in a different algorithm.

A vulnerability has been identified in SiNVR/SiVMS Video Server (All versions < V5.0.0). The two FTP services (default ports 21/tcp and 5411/tcp) of the SiVMS/SiNVR Video Server contain a path traversal vulnerability that could allow an authenticated remote attacker to access and download arbitrary files from the server, if the FTP services are enabled.

Versions of the npm CLI prior to 6.13.3 are vulnerable to an Arbitrary File Write. It fails to prevent access to folders outside of the intended node_modules folder through the bin field. A properly constructed entry in the package.json bin field would allow a package publisher to modify and/or gain access to arbitrary files on a user's system when the package is installed. This behavior is still possible through install scripts. This vulnerability bypasses a user using the --ignore-scripts install option.

Improper limitation of a pathname to a restricted directory vulnerabilities in FortiSandbox 3.2.0 through 3.2.2, and 3.1.0 through 3.1.4 may allow an authenticated user to obtain unauthorized access to files and data via specifially crafted web requests.

IBM Langflow OSS 1.0.0 through 1.10.0 allows an authenticated attacker to read arbitrary files including the JWT signing key and forge authentication tokens for any user.

Directory traversal vulnerability in OpenBase 10.0.5 and earlier allows remote authenticated users to create files with arbitrary contents via a .. (dot dot) in the first argument to the GlobalLog stored procedure. NOTE: this can be leveraged to execute arbitrary code using CVE-2007-5926.

Kedro is a toolbox for production-ready data science. Prior to 1.3.0, the _get_versioned_path() method in kedro/io/core.py constructs filesystem paths by directly interpolating user-supplied version strings without sanitization. Because version strings are used as path components, traversal sequences such as ../ are preserved and can escape the intended versioned dataset directory. This is reachable through multiple entry points: catalog.load(..., version=...), DataCatalog.from_config(..., load_versions=...), and the CLI via kedro run --load-versions=dataset:../../../secrets. An attacker who can influence the version string can force Kedro to load files from outside the intended version directory, enabling unauthorized file reads, data poisoning, or cross-tenant data access in shared environments. This vulnerability is fixed in 1.3.0.

WWBN AVideo is an open source video platform. In versions up to and including 26.0, the `objects/import.json.php` endpoint accepts a user-controlled `fileURI` POST parameter with only a regex check that the value ends in `.mp4`. Unlike `objects/listFiles.json.php`, which was hardened with a `realpath()` + directory prefix check to restrict paths to the `videos/` directory, `import.json.php` performs no directory restriction. This allows an authenticated user with upload permission to: (1) steal any other user's private video files by importing them into their own account, (2) read `.txt`/`.html`/`.htm` files adjacent to any `.mp4` file on the filesystem, and (3) delete `.mp4` and adjacent text files if writable by the web server process. Commit e110ff542acdd7e3b81bdd02b8402b9f6a61ad78 contains a patch.

Dagu is a workflow engine with a built-in Web user interface. From version 2.0.0 to before version 2.3.1, the fix for CVE-2026-27598 added ValidateDAGName to CreateNewDAG and rewrote generateFilePath to use filepath.Base. This patched the CREATE path. The remaining API endpoints - GET, DELETE, RENAME, EXECUTE - all pass the {fileName} URL path parameter to locateDAG without calling ValidateDAGName. %2F-encoded forward slashes in the {fileName} segment traverse outside the DAGs directory. This issue has been patched in version 2.3.1.

SciTokens is a reference library for generating and using SciTokens. Prior to version 1.9.7, the Enforcer is vulnerable to a path traversal attack where an attacker can use dot-dot (..) in the scope claim of a token to escape the intended directory restriction. This occurs because the library normalizes both the authorized path (from the token) and the requested path (from the application) before comparing them using startswith. This issue has been patched in version 1.9.7.

Budibase is a low code platform for creating internal tools, workflows, and admin panels. In 3.31.5 and earlier, a path traversal vulnerability in the PWA (Progressive Web App) ZIP processing endpoint (POST /api/pwa/process-zip) allows an authenticated user with builder privileges to read arbitrary files from the server filesystem, including /proc/1/environ which contains all environment variables — JWT secrets, database credentials, encryption keys, and API tokens. The server reads attacker-specified files via unsanitized path.join() with user-controlled input from icons.json inside the uploaded ZIP, then uploads the file contents to the object store (MinIO/S3) where they can be retrieved through signed URLs. This results in complete platform compromise as all cryptographic secrets and service credentials are exfiltrated in a single request.

SFTPGo is an open source, event-driven file transfer solution. In SFTPGo versions prior to 2.7.1, a path normalization discrepancy between the protocol handlers and the internal Virtual Filesystem routing can lead to an authorization bypass. An authenticated attacker can craft specific file paths to bypass folder-level permissions or escape the boundaries of a configured Virtual Folder. This vulnerability is fixed in 2.7.1.

A path traversal vulnerability was identified in GitHub Enterprise Server that allowed an attacker who had code execution inside the Dependabot updater container to write files to arbitrary repository paths, including GitHub Actions workflow files under .github/workflows/ as the path validation did not check the effective path which the attacker could control through the dependency file's directory and symlink target. If the repository used a pull_request_target workflow or had auto-merge enabled, an injected workflow could execute with access to the repository's GitHub Actions secrets. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.22 and was fixed in versions 3.21.3, 3.20.5, 3.19.9, 3.18.12, 3.17.18.

A vulnerability in a certain REST API endpoint of Cisco&nbsp;Data Center Network Manager (DCNM) Software could allow an authenticated, remote attacker to perform a path traversal attack on an affected device. The vulnerability is due to insufficient path restriction enforcement. An attacker could exploit this vulnerability by sending crafted HTTP requests to an affected device. A successful exploit could allow the attacker to overwrite or list arbitrary files on the affected device.Cisco has released software updates that address this vulnerability. There are no workarounds that address this vulnerability.

A vulnerability in the sfmgr daemon of Cisco Firepower Management Center (FMC) Software and Cisco Firepower Threat Defense (FTD) Software could allow an authenticated, remote attacker to perform directory traversal and access directories outside the restricted path. The vulnerability is due to insufficient input validation. An attacker could exploit this vulnerability by using a relative path in specific sfmgr commands. An exploit could allow the attacker to read or write arbitrary files on an sftunnel-connected peer device.

Pydio Cells 2.0.4 allows an authenticated user to write or overwrite existing files in another user’s personal and cells folders (repositories) by uploading a custom generated ZIP file and leveraging the file extraction feature present in the web application. The extracted files will be placed in the targeted user folders.

In drivers/target/target_core_xcopy.c in the Linux kernel before 5.10.7, insufficient identifier checking in the LIO SCSI target code can be used by remote attackers to read or write files via directory traversal in an XCOPY request, aka CID-2896c93811e3. For example, an attack can occur over a network if the attacker has access to one iSCSI LUN. The attacker gains control over file access because I/O operations are proxied via an attacker-selected backstore.

Problem in the Small HTTP Server v3.06.36 service. An authenticated path traversal vulnerability in '/' allows remote users to bypass the intended restrictions of SecurityManager and display any file if they have the appropriate permissions outside the document root configured on the server.

Roxy-WI is a web interface for managing Haproxy, Nginx, Apache and Keepalived servers. In versions 8.2.6.4 and prior, EscapedString (app/modules/roxywi/class_models.py:16-30) is the centralised Pydantic validator used on dozens of fields including SSH credential name, username, description, etc. Its if/elif/elif/else flow returns the metacharacter-stripped value without also enforcing the .. block. An attacker who appends a single ;, &, |, $, or backtick to a .. payload routes the value through the strip arm, where .. survives unblocked and the result is not shlex.quote()'d either. At time of publication, there are no publicly available patches.

Roxy-WI is a web interface for managing Haproxy, Nginx, Apache and Keepalived servers. In versions 8.2.6.4 and prior, ommit d4d10006 ("Expand validation to block .. in config_file_name and configver for improved security") added a line in app/modules/config/config.py:462. This is tuple-membership, not substring containment — '..' in (a, b, c) evaluates to True only if any of a, b, c is equal to the literal string '..'. For any realistic path-traversal payload (../../etc/passwd, ..\\..\\etc\\passwd, etc.) the check returns False and the patch silently lets the payload through. At time of publication, there are no publicly available patches.

The web-services interface of Loadbalancer.org Enterprise VA MAX through 8.3.8 could allow an authenticated, remote, low-privileged attacker to conduct directory traversal attacks and obtain read and write access to sensitive files.

A path traversal vulnerability in the Juniper Networks Junos OS device may allow an authenticated J-web user to read files with 'world' readable permission and delete files with 'world' writeable permission. This issue does not affect system files that can be accessed only by root user. This issue affects Juniper Networks Junos OS: 12.3 versions prior to 12.3R12-S13; 12.3X48 versions prior to 12.3X48-D85 on SRX Series; 14.1X53 versions prior to 14.1X53-D51; 15.1F6 versions prior to 15.1F6-S13; 15.1 versions prior to 15.1R7-S5; 15.1X49 versions prior to 15.1X49-D180 on SRX Series; 15.1X53 versions prior to 15.1X53-D238 on QFX5200/QFX5110 Series; 16.1 versions prior to 16.1R4-S13, 16.1R7-S5; 16.2 versions prior to 16.2R2-S10; 17.1 versions prior to 17.1R3-S1; 17.2 versions prior to 17.2R1-S9, 17.2R3-S2; 17.3 versions prior to 17.3R2-S5, 17.3R3-S5; 17.4 versions prior to 17.4R2-S9, 17.4R3; 18.1 versions prior to 18.1R3-S8; 18.2 versions prior to 18.2R3; 18.3 versions prior to 18.3R2-S3, 18.3R3; 18.4 versions prior to 18.4R2; 19.1 versions prior to 19.1R1-S4, 19.1R2.

An issue was discovered in Reprise License Manager 11.0. This vulnerability is a Path Traversal where the attacker, by changing a field in the Web Request, can have access to files on the File System of the Server. By specifying a pathname in the POST parameter "lf" to the goform/edit_lf_get_data URI, the attacker can retrieve the content of a file.

Dapr is a portable, event-driven, runtime for building distributed applications across cloud and edge. From versions 1.3.0 to before 1.15.14, 1.16.0-rc.1 to before 1.16.14, and 1.17.0-rc.1 to before 1.17.5, a vulnerability has been found in Dapr that allows bypassing access control policies for service invocation using reserved URL characters and path traversal sequences in method paths. The ACL normalized the method path independently from the dispatch layer, so the ACL evaluated one path while the target application received a different one. This issue has been patched in versions 1.15.14, 1.16.14, and 1.17.5.

Jellyfin is a free-software media system. Versions starting with 10.8.0 and prior to 10.8.10 and prior have a directory traversal vulnerability inside the `ClientLogController`, specifically `/ClientLog/Document`. When combined with a cross-site scripting vulnerability (CVE-2023-30627), this can result in file write and arbitrary code execution. Version 10.8.10 has a patch for this issue. There are no known workarounds.

Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal') vulnerability in Schweitzer Engineering Laboratories SEL-5036 acSELerator Bay Screen Builder Software on Windows allows Relative Path Traversal. SEL acSELerator Bay Screen Builder software is distributed by SEL-5033 SEL acSELerator RTAC, SEL-5030 Quickset, and SEL Compass. CVE-2023-31167 and was patched in the acSELerator Bay Screen Builder release available on 20230602. Please contact SEL for additional details. This issue affects SEL-5036 acSELerator Bay Screen Builder Software: before 1.0.49152.778.