CVE-2021-41038

The CVE-2021-41038 entry concerns the @theia/plugin-ext component of Eclipse Theia (pre-1.18.0). The issue is that Webview contents can be hijacked via postMessage(), caused by improper verification of the communication channel. This mode of exploitation could expose or modify Webview content dep...

CVE-2026-44688

The vulnerability CVE-2026-44688 affects Eclipse Theia versions prior to 1.71.0. The AI chat agent processes workspace file and directory names as part of its prompt context without distinguishing them from system instructions, enabling indirect prompt injection when an attacker uses adversarial ...

CVE-2026-60009

In Eclipse Theia up to 1.73.1, the backend @theia/filesystem binds POST /file-upload in every filesystem-enabled deployment. The handler uses an attacker-supplied absolute path from the multipart uri field and calls fs.move(tmp, target, { overwrite: true }) with no workspace confinement and no au...

CVE-2026-44691

CVE-2026-44691 affects Eclipse Theia versions before 1.69.0. The issue arises when custom task definitions in workspace files (e.g., .theia/tasks.json, .vscode/tasks.json) can be executed without workspace trust, potentially enabling arbitrary commands to run with the user’s privileges if a malic...

CVE-2026-61891

CVE-2026-61891 affects Eclipse Theia up to 1.73.1, specifically the @theia/filesystem backend. The HTTP endpoints GET /file, GET /files/, and PUT /files/ map a client-supplied URI directly to a filesystem path and stream the file, without confinement to the workspace or any allow-listed root. In ...

CVE-2026-46580

Theia before v1.71.0 loads files matching .prompts/*.prompttemplate from a workspace, allowing attacker-controlled content to override the AI agent’s system prompts (indirect prompt injection). This enables attack chains with untrusted workspaces, potentially causing data exfiltration via Markdow...

CVE-2026-22551

Eclipse Theia versions before 1.71.0 are affected: the AI chat could render Markdown image tags from AI responses, causing HTTP requests to arbitrary external URLs. In combination with a malicious workspace via prompt injection, an attacker could coax the AI agent to construct image URLs that lea...