ragleap-core v0.9.0 release notes

The biggest update since v0.7.0. RagLeap is now an open-source AI-employee platform you can install in minutes, with approval on every action.

New

AI Office dashboard (/office): overview, approvals, employees by department, task board, agent runs, activity log, and a Settings tab. Pick the AI provider and models, paste keys (encrypted, never shown again) and test the connection.
Settings in the dashboard. A value saved there wins over .env, and the page shows where each value comes from.
Embeddings are no longer Gemini-only. Use Ollama, OpenAI, Mistral and more. The vector size is checked at startup, and tables that already hold embeddings are never changed.
Installer and ragleap command. The installer generates your keys, offers Gemini, local Ollama or skip, and prints a one-click dashboard link. Commands: launch, stop, status, logs, update (backs up the database first) and key.
Agents. Act-observe agent loop, sandboxed code and shell, read-only page fetch, MCP client, task tickets, and scheduled triggers with cron and time zones.
Approval inbox. Approve or reject actions in the dashboard or over the API.

Approval modes (off, semi, full), sensitive roles always need approval, a taint rule after web content, a network-less sandbox, and SSRF-safe page fetching.
API-key protection on all data routes, and no exception text returned to callers.

The installer is tested offline and has not yet had a real clean-machine run.
Local models on a CPU are slower and less accurate than Gemini.
Public access, login throttling, an interactive browser and more channels are not built yet.
The dashboard is reachable through an SSH tunnel or locally, by design.

Run ragleap update, or git pull and docker compose up --build -d. The database changes are additive.