pg-aiguide
pg-aiguide is an MCP server and Claude plugin from Timescale that provides AI coding assistants with semantic search over PostgreSQL documentation and curated best-practice skills, producing more accurate, modern, and constraint-rich SQL and schema designs.
Score Breakdown
⚙ Agent Friendliness
🔒 Security
PostgreSQL AI guide MCP — wraps pg with AI assistance. Connection string contains credentials — use env vars or secrets manager. Use read-only DB user for query-only agents.
⚡ Reliability
Best When
An engineer is using an AI coding assistant to design or modify PostgreSQL schemas and wants production-quality output with proper indexing, constraints, and modern PG features rather than generic boilerplate.
Avoid When
The database is not PostgreSQL, or the team has its own internal schema standards that override general best practices.
Use Cases
- • Generating PostgreSQL schemas with proper constraints, indexes, and PG17-recommended patterns instead of generic SQL
- • Searching version-aware PostgreSQL documentation during schema design or query optimization tasks
- • Getting TimescaleDB and PostGIS documentation and best practices surfaced automatically in AI coding contexts
- • Reducing AI hallucination of outdated or incorrect PostgreSQL patterns through curated skill injection
Not For
- • Non-PostgreSQL databases (MySQL, SQLite, MongoDB, etc.)
- • Developers who don't use AI coding assistants
- • Simple CRUD apps where schema quality is not a priority
Interface
Authentication
No authentication required for the public MCP server. One-click installs for Cursor, VS Code, and others.
Pricing
Apache 2.0 license. Public MCP server at mcp.tigerdata.com is free to use.
Agent Metadata
Known Gotchas
- ⚠ Public server at mcp.tigerdata.com depends on Timescale infrastructure availability
- ⚠ pgvector documentation support listed as coming soon
- ⚠ Skills are opinionated Timescale best practices; may conflict with organization-specific conventions
- ⚠ Version-aware docs mean agents should pass Postgres version when querying for accurate results
Alternatives
Full Evaluation Report
Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for pg-aiguide.
Scores are editorial opinions as of 2026-03-06.