Signal is active with 3 mentions in latest observed window
Recommended action window: Monitor and triage in normal cycle
NVD description
OpenHuman is an open-source agent harness available as a desktop application and Rust library. Before 0.57.13, a remote direct message received through a configured Yuanbao channel is converted into a generic channel AgentTurnRequest that uses the default tool policy without establishing an interactive approval context. Under the default Supervised autonomy policy with Node enabled, node_exec is classified as a write operation that should prompt for approval, but ApprovalGate treats missing APPROVAL_CHAT_CONTEXT as a non-interactive turn and allows execution. Because Yuanbao direct messages are open by default, an external sender who can message the connected bot can influence the agent to execute JavaScript and operating-system commands as the OpenHuman user without an approval prompt. The issue requires a configured Yuanbao channel, the default approval-requiring policy state, Node availability, and successful model selection of node_exec. This issue is fixed in version 0.57.13.
Priority
LOW
Exploitation
NONE
PoC
NONE
Patch
NONE
Momentum
STABLE
🆕 New CVEs tinyhumansai openhuman: 2 critical CVEs Worst: remote command execution via inbound channel cron jobs (CVE-2026-54143) Patch → 0.57.13 https://cve.now/CVE-2026-54143?release=CVE-2026-54141&utm_source=x&utm_medium=social&utm_campaign=cvedotnow&utm_content=rollup
🚨Critical - OpenHuman Approval Bypass RCE (CVE-2026-54141, CVE-2026-54143) Remote messages to configured OpenHuman channels can bypass human approval, enabling JavaScript/OS command execution through Yuanbao DMs and node_exec, or persistent shell execution through unapproved scheduled jobs. Commands run with the OpenHuman user’s privileges; the demonstrated scheduled-job exploit requires an existing helper executable. 👉Affected: openhuman < 0.57.13 | Upgrade to 0.57.13
OpenHuman from @tinyhumansai patched CVE-2026-54141 (CVSS 9.0). A Yuanbao DM had no approval context, so the gate treated it as non-interactive and let node_exec run. No prompt. Agent builders: missing approval context must fail closed. Update to 0.57.13.
@tinyhumansai CVE details: https://critical.is/cve/CVE-2026-54141 Sibling bug (Slack, email and Notion content injecting into background agents): https://secopssignal.com/news/vulnerabilities/openhuman-injection-flaw-lets-slack-email-trigger/
🔴 OpenHuman: Yuanbao non-web channel can bypass human approval to execute node_exec (CVE-2026-54141) Fix → 0.57.13 https://agentprime.io/advisories/CVE-2026-54141/
OpenHuman before 0.57.13 has two critical flaws in its approval gate. CVE-2026-54143 (CVSS 9.8): a remote chat message can create an unapproved shell cron job. CVE-2026-54141 (CVSS 9.0): a Yuanbao DM can trigger node_exec without approval. VulnTracker recommends updating to 0.57.13 and reviewing cron jobs for unknown shell jobs. Details: http://vulntracker.io/cves/CVE-2026-54143 #OpenHuman #CVE #VulnTracker