CVE Vulnerability Catalog

In PraisonAI prior to 0.1.6, when PLATFORM_JWT_SECRET is unset, the public dev-secret-change-me HS256 signing key is used, and startup and token-issuance guards are disabled because PLATFORM_ENV defaults to dev. An unauthenticated attacker can sign a JWT with an arbitrary sub value that is accepted as an authenticated identity, enabling impersonation of a user or workspace owner.

In PraisonAI prior to 0.1.6, when PLATFORM_JWT_SECRET is unset, JWT_SECRET is assigned the public dev-secret-change-me value, and its production guard does not run when PLATFORM_ENV is also unset (defaults to dev). A remote unauthenticated attacker can mint an HS256 token with arbitrary sub and email, which is accepted by AuthService._verify_token() and get_current_user for protected API routes.

In PraisonAI prior to 1.7.2, the codeMode tool in src/praisonai-ts/src/tools/builtins/code-mode.ts executes model-generated JavaScript with new Function() and with(sandbox), while a regex blocklist can be bypassed with Function('return this')() to recover the global object and by dynamically constructing the child_process module name. An attacker who can influence the code argument can access host process capabilities, read or write files, obtain environment credentials, and execute OS commands with PraisonAI process privileges.

In PraisonAI from 1.6.0 until 1.7.2, AgentOS in src/praisonai-ts/src/os/agentos.ts uses the 0.0.0.0 default from src/praisonai-ts/src/os/config.ts and registers GET /api/agents and POST /api/chat without authentication middleware. A remote caller who can reach the service can obtain agent names, roles, and instruction prefixes and can invoke a selected agent, potentially reaching its tools, memory, external APIs, credentials, and workflow state.

In PraisonAI from 1.5.0 until 1.7.2, MCPServer.startHttp() in src/praisonai-ts/src/mcp/server.ts binds without a host restriction and forwards every HTTP POST request to handleRequest() without authentication or authorization. Any network client that can reach the port can call tools/list, tools/call, resources/read, or prompts/get, causing registered handlers to run with server-side credentials and process privileges or disclose registered data.

In PraisonAI from 1.4.0 until 1.7.2, codeMode in src/praisonai-ts/src/tools/builtins/code-mode.ts executes untrusted JavaScript with new Function() inside with(sandbox) and relies on a small source-code blocklist plus shadowed process and require properties. Code can use ({}).constructor.constructor to recover the real Function constructor, obtain process and process.mainModule.require, and reach host filesystem and subprocess APIs despite the advertised sandbox. Attackers who control codeMode input can read secrets, modify files, execute commands, or exhaust the host process.

PraisonAI is a multi-agent teams system. From 1.4.0 until 1.7.2, createAgentLoop() in src/praisonai-ts/src/ai/agent-loop.ts passes executable tools to generateText() before invoking the onToolCall approval callback. Because the wrapped AI SDK executes tool handlers during generation, a callback that returns false records tool_rejected only after the denied tool has already produced side effects.

PraisonAI is a multi-agent teams system. From 1.2.3 until 1.7.2, CommandValidator in src/praisonai-ts/src/cli/features/sandbox-executor.ts validates only the first whitespace-delimited executable against allowedCommands, then SandboxExecutor passes the complete command string to sh -c. A command beginning with an allowed executable can append a non-allowlisted command through shell metacharacters.

PraisonAI is a multi-agent teams system. From 1.2.3 until 1.7.2, SandboxExecutor network-isolated mode in src/praisonai-ts/src/cli/features/sandbox-executor.ts uses buildEnv() only to inject invalid http_proxy and https_proxy environment variables and does not establish an OS-level network boundary. Programs that ignore those proxy variables can open sockets directly.

PraisonAI is a multi-agent teams system. From 1.5.1 until 1.7.2, MCPSecurity.evaluatePolicy() in src/praisonai-ts/src/mcp/security.ts invokes the configured credential validator only when AuthMethod is api-key or bearer. Basic and OAuth policies accept any non-empty Authorization header without calling auth.validate(), then return an authenticated result.

PraisonAI is a multi-agent teams system. From 1.5.1 until 1.7.2, the shell() helper exported from src/praisonai-ts/src/tools/utility-tools.ts checks only the first whitespace-delimited token against safeCommands and then passes the complete original string to child_process.exec(). A string that starts with an allowed read-only command can append a second non-allowlisted command through shell syntax.

In PraisonAI, ToolsMCPServer.run_sse() in src/praisonai-agents/praisonaiagents/mcp/mcp_server.py mounts SseServerTransport on the legacy /sse and /messages/ endpoints without default Host, Origin, or authentication enforcement. A malicious website can use DNS rebinding against a reachable local or internal SSE server, supply attacker-controlled Host and Origin headers, enumerate registered tools, and invoke them with the server user's privileges. An initial remediation was released in praisonaiagents 1.6.59 and PraisonAI 4.6.59.

Kimai is an open-source time tracking application. Prior to 2.58.0, ExportController::createExportTemplate() and ExportController::editExportTemplate() inherit only the class-level create_export permission, which ROLE_TEAMLEAD receives by default, and omit the create_export_template permission required by the API routes and user interface. A teamlead can directly access the export template creation and editing web routes to create or modify global ExportTemplate records marked available to all users, altering export columns, renderer, format, and output used by other users and administrators. This issue is fixed in version 2.58.0.

In Kimai, the KIMAI_SESSION cookie issued after password verification but before TOTP completion is accepted by every /api route because config/packages/security.yaml protects the API with IS_AUTHENTICATED and App\API\Authentication\ApiRequestMatcher routes an existing session through the main firewall. A Scheb TwoFactorToken satisfies that access rule, and App\Voter\ApiVoter grants API access to its User, allowing an attacker with a valid account password to use authenticated REST API operations without entering the second factor even though web routes remain blocked. This issue is fixed in version 2.59.0.

Kimai is an open-source time tracking application. Prior to 2.57.0, GET or POST requests to /en/admin/project/{id}/rate/{rate}, /en/admin/customer/{id}/rate/{rate}, and /en/admin/activity/{id}/rate/{rate} independently resolve the authorized parent identifier and the attacker-selected child rate identifier without confirming that the ProjectRate, CustomerRate, or ActivityRate belongs to that parent. An authenticated user who can edit one parent object can pair it with a rate record from an unauthorized project, customer, or activity and persist changes to billing configuration in kimai2_projects_rates, kimai2_customers_rates, or kimai2_activities_rates. This issue is fixed in version 2.57.0.

Kimai is an open-source time tracking application. Prior to 2.58.0, POST /api/teams/{id}/members/{userId} and POST /api/teams/{id}/activities/{activityId} verify that a teamlead may edit the Team but do not verify access_user for the referenced User or view access for the referenced Activity. A teamlead can add users or activities outside the teamlead's manageable scope to an editable team, bypassing the narrower choices enforced by TeamEditForm and UserRepository::getQueryBuilderForFormType(). The resulting relationships can be trusted by RolePermissionManager::checkTeamAccessActivity() and other team-based authorization, visibility, reporting, and workflow logic. This issue is fixed in version 2.58.0.

In Kimai prior to 2.58.0, the official Docker image sets APP_SECRET to the public value change_this_to_something_unique in Dockerfile, and .docker/entrypoint.sh neither replaces nor rejects that value before Symfony uses it as kernel.secret. An unauthenticated attacker who reaches a deployment that did not override APP_SECRET, knows a username, correctly guesses the account ID associated with that username, and targets an account without active two-factor authentication can forge HMAC-protected authentication artifacts, including KIMAI_REMEMBER cookies and login links, to access the account without its password. The updated entrypoint generates and persists a random secret when no safe operator-provided value exists.

Kimai is an open-source time tracking application. Prior to 2.58.0, TimesheetController exposes GET /api/timesheets/{id}/stop and GET /api/timesheets/{id}/restart, which reuse an authenticated browser session and perform state-changing operations through GET requests without a request-forgery defense. A remote attacker can cause a logged-in user to request either route from attacker-controlled content, stopping a running timesheet or restarting a historical entry to create and start a new timesheet without the user's consent. The unauthorized changes can corrupt time records, billing, reports, approvals, and audits. This issue is fixed in version 2.58.0.

Kimai is an open-source time tracking application. Prior to 2.58.0, PATCH /api/timesheets/{id}/restart, PATCH /api/timesheets/{id}/duplicate, and the web duplicate workflow can derive a new record from an owned historical timesheet after the user's access to its project or activity has been revoked. TimesheetVoter evaluates the own-timesheet permission before current team access, and its canStart() logic validates object visibility but does not verify the user's current team access to the referenced project and activity. An old entry therefore acts as a persistent capability to create new time records under an unauthorized project and activity, corrupting budgets, statistics, reports, and invoices after an administrative revocation. This issue is fixed in version 2.58.0.

Prior to version 2.57.0 of Kimai, GET or POST requests to /en/admin/activity/create/{project} and /en/admin/project/create/{customer} required only the generic create_activity or create_project capability and did not verify edit access to the supplied Project or Customer object. A user knowing a valid project.id or customer identifier could create a new child business object under an unauthorized parent, polluting project or customer configuration and influencing later time-entry, rate, reporting, and billing behavior.