漏洞公告

OpenClaw: Feishu dynamic-agent bindings could miss configWrites enforcement

OpenClaw: Sandboxed session spawn could expose the real workspace path to child prompts

OpenClaw: Embedded runner policy could be confused by provider aliases

OpenClaw: Fake package roots could influence memory-core artifact loading

OpenClaw: Workspace .env could override Homebrew executable selection for skill install flows

OpenClaw: QQBot pre-dispatch slash commands could skip allowFrom checks

OpenClaw: Non-owner chat senders could issue device-pairing bootstrap codes

OpenClaw: Browser debug/export routes could reuse already-open blocked tabs

OpenClaw: message.action forwarding could send Gateway credentials to model-supplied loopback URLs

OpenClaw: QQBot admin commands could skip DM-only and allowFrom policy

OpenClaw: Mattermost handlers could fall open when channel type was missing

OpenClaw: Trusted-proxy Control UI WebSocket accepted client-declared scopes before pairing

OpenClaw: Slack allowFrom could bind to mutable display names

OpenClaw: Skill Workshop apply flow could override pending approval

OpenClaw: QQBot streaming command could mutate config without explicit allowFrom

OpenClaw: Node pairing reconnection could confuse approval scope state

OpenClaw's Slack plugin approvals used the exec approver gate for plugin actions

OpenClaw: memory-wiki ingest could read local files with operator.write scope

OpenClaw: Scoped chat.send route inheritance could bypass admin command scope gates

OpenClaw's POSIX node system.run safe-bin allowlist could be widened by shell expansion

OpenClaw: QQBot native approval buttons did not enforce configured approver identity

OpenClaw: Hook-triggered CLI runs could receive owner MCP tool authority

OpenClaw: Control UI locality spoofing could mint a durable admin device token