Loading...
Loading...
Found 5 Skills
Use the Polygres Python SDK to write individual standard-project application records and build retrieval on standard or synchronized projects with Polygres AI Context, graph, text, hybrid, and existing vector queries; manage supported pgContext collections, points, and durable operations; configure safe Runtime API clients; paginate typed results; and handle SDK errors. Use for Python backend code calling a project's Runtime API. Do not use for sync control-plane administration, bulk imports, migrations, or interactive CLI workflows.
Design and review Polygres retrieval plans across relational, graph, text, hybrid, Polygres AI Context, and existing pgvector-backed strategies. Use when a user must choose between retrieval surfaces, model graph or embedding inputs, plan Context collections or grounded RAG, migrate or coexist with an existing vector configuration, or review retrieval architecture before configuration or implementation. Do not mutate a project, run configuration commands, or write application code.
Diagnose Polygres CLI, Runtime API, control-plane, standard or synchronized PostgreSQL project, job, migration, retrieval, and pgContext failures using public read-only evidence. Use for errors, sync preflight or lifecycle failures, timeouts, partial failures, readiness problems, ambiguous projects, failed Context collections or operations, stale point mappings, or broken graph, vector, text, hybrid, or Joint retrieval. Do not use private observability, undocumented endpoints, or mutating repair actions.
Set up or extend Polygres from either a short request such as "Help me set up Polygres" or a detailed ingestion, memory, graph, embedding, synchronization, or retrieval specification. Also use for questions such as "What can I do with Polygres?" by scanning the accessible current workspace and project read-only and giving a personalized recommendation without changing anything. Ask one short direction question first when a setup request identifies neither a source nor an outcome; otherwise inspect the user's accessible data and application, resolve only critical unknowns, design the smallest useful schema and retrieval setup, generate source-specific ingestion and retrieval code, configure the selected project after one consolidated approval, verify a small vertical slice, and optionally connect capture and recall to the user's agent. Use whenever the user intends to make their data usable through Polygres, even if they do not say "data pipeline."
Use the Polygres CLI to authenticate, create standard or synchronized PostgreSQL projects, select and inspect projects, write one standard-project record through the public Runtime rows surface, import data, apply migrations, configure graph, text, or Polygres AI Context retrieval, manage Runtime keys and durable operations, and check readiness. Use for documented managed-project operations through public CLI workflows. Use polygres-troubleshooting for evidence-first diagnosis of an unexplained failure.