Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
Combines opinionated Postgres skills with version-aware search of the official manual. The MCP route adds documentation retrieval, while the skill packages provide reusable guidance for schema and query work.
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Follow the documented setup, then try a first task.
I’m using
Claude Code pg plugin
Installs: The complete pg@aiguide plugin, including the repository's skills bundle and the public PostgreSQL documentation MCP server. The listed skill paths are a bounded selection. · Claude Code
Before you start
Claude Code available in the terminal.
Run both commands in your terminal, in order, to add the marketplace and install its pg plugin.
claude plugin marketplace add timescale/pg-aiguide
claude plugin install pg@aiguide
Installs: Connect Codex to the public pg-aiguide MCP server. The supplied connection instructions do not establish individual skill-file mappings. · Codex
Before you start
Codex available in the terminal.
Run this command in your terminal to add the MCP server to Codex.
Installs: Connect Claude Code to the public pg-aiguide MCP server. The supplied connection instructions do not establish individual skill-file mappings. · Claude Code
Before you start
Claude Code available in the terminal.
Run this command in your terminal to connect Claude Code to the MCP server.
claude mcp add --transport http pg-aiguide https://mcp.tigerdata.com/docs
Required user and order fields, business constraints, and common queries.
Use the design-postgres-tables skill to draft PostgreSQL users and orders tables for a small application. Use my supplied fields and access patterns to choose data types, primary and foreign keys, required constraints, and indexes. Explain the main choices and return SQL for review without executing it.
One business concept or focused relationship question.
Provided schema files or an available authorized PostgreSQL connection.
Authorized database and schema scope.
Use the schema-exploration skill to identify the tables representing one business concept I specify and explain their relevant keys and relationships. Work from my provided schema files or the authorized database connection, using structural catalog metadata only. Report schema-qualified objects, distinguish verified facts from inferred meaning, and stop once this question is answered.
Embedding dimension and the model used consistently for stored and query embeddings.
Document fields to retain alongside embeddings.
Use the pgvector-semantic-search skill to draft a minimal setup for storing document text with embeddings and retrieving the five nearest matches. Use the default halfvec storage, cosine distance, and HNSW configuration, adapting every dimension to my embedding dimension. Include a parameterized query with an explicit vector cast and briefly explain the index and recall settings. Return SQL for review without executing it.