sql-firewall

PostgreSQL/MySQL wire-protocol proxy for database guardrails for AI agents: stops destructive SQL before it reaches your database, no code changes. Source-available (FSL-1.1-ALv2, Apache-2.0 after two years).

Vericto CLI, database guardrails for AI agents in your terminal: check SQL against your Vericto rules before it runs, in CI, hooks or offline. brew install vericto/tap/vericto · npm i -g @vericto/vericto-cli. Source-available (FSL-1.1-ALv2).

Official Vericto SDK for Python, database guardrails for AI agents: check SQL before it runs (Snowflake, SQLite, SQLAlchemy, LangChain, LlamaIndex, dbt, Airflow). Apache-2.0.

Official Vericto SDK for Node.js, database guardrails for AI agents: check SQL before it runs (Snowflake, better-sqlite3, LangChain.js). Apache-2.0.

Deterministic SQL AST engine behind Vericto's database guardrails for AI agents: parses each statement and evaluates it against your rules, so destructive queries are blocked before they run. Source-available (FSL-1.1-ALv2, Apache-2.0 after two years).

Homebrew tap for the Vericto CLI: brew install vericto/tap/vericto

Scoop bucket for the Vericto CLI: scoop bucket add vericto https://github.com/vericto/scoop-bucket

SQL firewall for AI agents: read-only execution, table/column allowlists, PII masking and tamper-evident audit for PostgreSQL, MySQL and SQLite. Pre-alpha.

The SQL firewall for AI agents: a default-deny sidecar / MCP enforcement point that governs what an AI agent runs against your database. Open core, AGPL-3.0.

Reproducible benchmark of vericto-proxy: open-loop latency, CPU and per-byte cost against direct, HAProxy, PgBouncer and ProxySQL (Apache-2.0)

Encrypt PostgreSQL fields at the application layer in Rust to protect data from DB access, leaked creds, and exposed backups