CVE-2026-27001
HighSummary
OpenClaw to osobisty asystent AI, który przed wersją 2026.2.15 wbudowywał bieżący katalog roboczy w systemowy prompt agenta bez odpowiedniej sanitizacji. Atakujący mógł wykorzystać to, aby wstrzyknąć kontrolowane przez siebie instrukcje, jeśli OpenClaw działał w katalogu zawierającym znaki kontrolne/formatujące.
Risk Assessment
Organizacja może być narażona na ataki, które wykorzystują wstrzyknięcie złośliwych instrukcji do systemu, co może prowadzić do nieautoryzowanego dostępu lub wykonania niebezpiecznych operacji.
Recommendation
Zaleca się aktualizację OpenClaw do wersji 2026.2.15 lub nowszej, aby zapewnić odpowiednią sanitizację ścieżek katalogów roboczych i zminimalizować ryzyko wstrzyknięcia złośliwego kodu.
Related vulnerabilities
- CVE-2026-87701Critical
Improper neutralization of special elements in output used by a downstream component ('injection') in Azure Cosmos DB allows an authorized attacker to elevate privileges over a network.
- CVE-2026-85889Critical
Missing authentication for a critical function in Azure AI Foundry allows an unauthorized attacker to elevate privileges over a network.
- CVE-2026-85885Critical
Improper neutralization of special elements used in a command ('command injection') in M365 Copilot allows an authorized attacker to elevate privileges over a network.
- CVE-2026-83944Critical
Improper access control in Azure Logic Apps allows an unauthorized attacker to elevate privileges over a network.
- CVE-2026-77903Critical
Authentication bypass by spoofing in Microsoft Dataverse allows an unauthorized attacker to elevate privileges over a network.
- CVE-2026-70200Critical
Improper limitation of a pathname to a restricted directory ('path traversal') in Azure Logic Apps allows an unauthorized attacker to elevate privileges over a network.
- CVE-2026-70009Critical
Improper limitation of a pathname to a restricted directory ('path traversal') in Azure Arc allows an unauthorized attacker to elevate privileges over a network.
- CVE-2026-69865Critical
Authorization bypass through a user-controlled key in Microsoft Container Registry allows an unauthorized attacker to elevate privileges over a network.
- CVE-2026-69399Critical
Elevation of Privilege vulnerability in Azure Arc. Technical details were not provided in the description.
- CVE-2026-76949Critical
A vulnerability in the ash_authentication library (Elixir) allows an attacker who can plant a remember-me cookie in a victim's browser to replace the victim's authenticated session with one for the attacker's own account. The issue stems from a mismatch between the session keys used by the guard and the reader, causing remember-me sign-in to run on every request.
Original NVD description (English source)
OpenClaw is a personal AI assistant. Prior to version 2026.2.15, OpenClaw embedded the current working directory (workspace path) into the agent system prompt without sanitization. If an attacker can cause OpenClaw to run inside a directory whose name contains control/format characters (for example newlines or Unicode bidi/zero-width markers), those characters could break the prompt structure and inject attacker-controlled instructions. Starting in version 2026.2.15, the workspace path is sanitized before it is embedded into any LLM prompt output, stripping Unicode control/format characters and explicit line/paragraph separators. Workspace path resolution also applies the same sanitization as defense-in-depth.