v1.9 closes two important security issues, adds the built-in A0 CLI Connector and its setup skill, restores lightweight lexical skill recall, gives Browser Agent runs their own model preset, and redesigns messaging integration setup flows across email, Telegram, and WhatsApp.
SSRF blocked in document_query remote fetching — Remote document fetching now validates URLs before any network request, blocks localhost and non-public IP targets, validates redirect hops, disables implicit proxy trust, and enforces a strict size cap. Third-party loaders no longer receive attacker-controlled URLs directly; content is prefetched and parsed from trusted local bytes instead. A follow-up compatibility fix also restores access to public sites that rejected the new request fingerprint.
Path traversal blocked in download_work_dir_file — Download requests are now rejected if their resolved path escapes the runtime base directory, preventing arbitrary file reads outside the allowed workspace.
✨ New Features
Built-in A0 CLI Connector plugin — Agent Zero now includes a host-side connector plugin so the A0 CLI can connect over authenticated HTTP and WebSocket, with capability discovery, chat/context lifecycle endpoints, log streaming, remote editing, code execution, and file-tree bridging.
a0-setup-cli built-in skill — A new setup skill guides users through host-side A0 connector installation with installer-first guidance, container-aware guardrails, fallback install paths, and updated Flare Tunnel connection guidance.
Restored lexical trigger-based skill matching — Lightweight trigger-word scoring is back in search_skills(), re-enabling skills_tool:search and lexical relevant-skill recall for the current user message without requiring vector-database recall.
Native chat controls for messaging integrations — Telegram, WhatsApp, and email threads now share transport-level commands like /project, /config, /send, and /queue send so you can manage the active chat directly from inside each integration.
Browser Agent model preset selection — Browser Agent runs can now use a dedicated _model_config preset instead of always inheriting the main model configuration.
🎨 UI & UX Improvements
Redesigned messaging integration settings — Email, Telegram, and WhatsApp settings panels now use clearer step-based setup flows, guided first-run experiences, provider presets for email, safer access warnings, richer test feedback, and responsive layouts. Advanced email routing, server, and scheduling options now live behind an Advanced section.
Componentized model config — The model configuration UI has been refactored into components, with the store split into mixins and API key management unified into one flow.
⚡ Other Improvements
Clearer plugin skill lifecycle guidance — Plugin skill documentation now formalizes install(), uninstall(), and preupdate() requirements when dependencies are involved.
Contributor sharing and fork safety docs — Added documentation to make collaboration and fork-based contribution workflows safer and easier to follow.
We use essential cookies for login and saving your preferences. With your consent, we also
use analytics cookies (Google Analytics) to improve the website.
Learn more