We're excited to announce pgconsole 1.0 — a self-hosted PostgreSQL editor designed for teams that care about security, simplicity, and speed.
Why pgconsole?
Most database tools fall into two camps: heavyweight platforms that require their own infrastructure, or lightweight editors that lack access control. PostgreSQL-specific editors exist but feel outdated, while modern editors spread thin across dozens of databases. pgconsole is a modern editor built exclusively for PostgreSQL — a single binary with everything you need.
By targeting PostgreSQL alone, we optimize every layer of the experience: native database/schema/table navigation, a real PostgreSQL parser powering autocomplete, syntax highlighting, code folding, and formatting — not generic SQL heuristics, but accurate PostgreSQL semantics.
GitOps Native
Everything lives in pgconsole.toml. Connections, users, groups, access rules — all in one file. No database migrations, no admin panels. Review access changes in pull requests the same way you review code.
[[connections]]
id = "production"
name = "Production"
host = "prod.example.com"
port = 5432
database = "myapp"
username = "app_user"
password = "prod_password"
ssl_mode = "require"
[[auth.providers]]
type = "google"
client_id = "..."
client_secret = "..."
Built-in Access Control
Seven permission levels — from explain (view query plans only) to admin (role and database management). Permissions are disjoint, not hierarchical, so you can grant exactly the access each team member needs.
[[iam]]
connection = "production"
permissions = ["read", "explain"]
members = ["group:engineering"]
[[iam]]
connection = "production"
permissions = ["*"]
members = ["group:dba"]
AI Assistant
Bring your own LLM — OpenAI, Anthropic Claude, or Google Gemini. pgconsole uses your database schema to generate, explain, fix, and rewrite SQL directly in the editor. It also evaluates change risk before you hit run.
[[ai.providers]]
id = "claude"
name = "Claude Sonnet"
vendor = "anthropic"
model = "claude-sonnet-4-20250514"
api_key = "sk-ant-..."
Audit Logging
Every query is logged with the user, connection, timestamp, and full SQL. No more guessing who ran that DELETE on production. Logs are emitted as JSON to stdout, so you can stream them directly to your SIEM of choice — Splunk, ELK, or any log aggregator.
Open Source
Connecting to your database is sensitive — so the entire codebase is public and available for inspection. pgconsole doesn't phone home. The only outbound connections are to your databases and, optionally, your configured AI provider.
Get Started
Install and start querying in under a minute. See the quickstart guide for full details.
npx @pgplex/pgconsole@latest
docker run -p 9876:9876 pgplex/pgconsole