Quick links

pg_llm_helper

Captures PostgreSQL errors in shared memory and exposes history plus an optional pgai-backed LLM troubleshooting function.

plpgsql	PL/pgSQL procedural language
vector	vector data type and ivfflat and hnsw access methods

pgmnemo	Single-plan multimodal agent memory for PostgreSQL
pg_llm	Integrates local and remote large-language models with chat, sessions, text-to-SQL, audit and vector-backed knowledge features.
pgclaw	Store LLM or OpenClaw agents as PostgreSQL values and process watched rows with a background worker
pgtoken	PostgreSQL extension for compact, codebook-compressed LLM token ID storage. Encode once, retrieve forever. O(1) token count without decoding.
verisql	Audit-grade verification gateway for AI-generated SQL — catches silently wrong queries before they hit your dashboard, keeps tamper-evident audit trails, verifies migration equivalence. Zero LLM tokens for 90% of checks.
pg_deeplake	Deeplake is AI Data Runtime for Agents. It provides serverless postgres with a multimodal datalake, enabling scalable retrieval and training.
vectorize	The simplest way to do vector search on Postgres
lantern_extras	Convenience functions for working with vector embeddings
rds_ai	Alibaba Cloud RDS extension for calling LLMs, generating embeddings, semantic retrieval, and end-to-end RAG from SQL
tencentdb_ai	TencentDB extension for calling large-model APIs and building in-database RAG workflows.