Malicious code in chatcc-agent (npm)
Historical advisory saved 24 Oct 2026 13:17:17 UTC. The publisher may have updated it since then. Latest advisory.
osv · MAL-2026-13259
Published--- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (8a69e686f3f793120840abd296a10a4b1c8d734797607e7dce82a82e90cf7cf0) chatcc-agent runs a daemon that connects to Tencent Cloud IM as a WebSocket client and treats inbound peer messages of type 'terminal_command', 'terminal_input', and 'terminal_pty_*' as instructions to spawn or write to a local shell / PTY on the installer's machine. In src/cclink-studio-remote.js and related handlers, _handleAdvancedCommand calls pty.spawn(shell, ['-lc', command],...) with the command string supplied by the remote IM peer, and handleInput forwards peer-supplied bytes into the active process's stdin. The result is a network-reachable, remote-controlled command-execution channel into the host that installed and runs the agent. Access is gated by JWT / paired-client authentication, but the channel, its bindings to a login shell, and the ability to stream arbitrary stdin exist by design and constitute a remote-shell surface regardless of how the package is described. Separately, the npm install script only performs a local native build (chmod of the shipped node-pty spawn-helper and compilation of a shipped C source with cc) and does not fetch or execute remote content.
- Source published
- 5 Aug 2026 15:38:53 UTC
- Source modified
- 5 Aug 2026 16:50:20 UTC
- Saved here
- 24 Oct 2026 13:17:17 UTC
- Source terms
- OSV · upstream source license applies
Original source record · Cite this exact observation · Complete original publisher JSON
Source: osv · OSV · upstream source license applies. The publisher's wording is preserved. ChangeIntel groups the information and labels changes. The JSON record includes additional source fields.