Плагин data-and-ai-tools

Build a bounded RAG feature against an accepted corpus, access contract, and measured retrieval and answer evaluation.

Design a source-grounded data model, access boundaries, indexes and a migration proposal without executing database changes.

Make this Claude Code desktop session TARS, the manager. Spawns chips for what you ask, manages every session to done, fields their questions, uses CASE for proportion and scoped decisions, and KIPP as the Feature Desk, runs a real 10-minute tick and a 1-minute CI watch. Does no work itself. Trigger

The vibe-coder's AI coding tool advisor and bounded coding task brief writer, recommends, compares, and configures Cursor, Claude Code, Aider, Cline, Windsurf (Cascade), Continue.dev, Replit Agent, Devin 2.0, and Bolt across four autonomy tiers. Invoke when the user says \"which AI coding tool shoul

The vibe coder's AI toolbox specialist: AI gateways (Portkey, OpenRouter), cloud providers (AWS Bedrock, Vertex AI, Azure OpenAI), frontier model selection (Claude, GPT, Gemini), cheap-fallback routes (Haiku, Mini, Flash), local LLMs (Ollama, LM Studio), GPU cloud (Runpod, Modal, Together, Fireworks

Bifrost gateway implementation drone for maximhq/bifrost trees - freeze and upgrade execution, admin API changes, plugin wiring, semantic cache config, contract extraction from docs/openapi. Use for any Bifrost coding task.

Cursor IDE platform specialist: project rules (.cursorrules migration, .cursor/rules/*.mdc authoring), MCP server registration and tool authoring, @cursor/sdk API for programmatic agent automation, custom modes, Agents Window and Cloud Agents, and Cursor productivity patterns. Invoke when the user s

PostgreSQL data architecture specialist: schema design, indexing strategy, zero-downtime migrations, ORM choice (Drizzle / Prisma / raw SQL), and serverless DB platform selection (Supabase / Neon / Turso / PlanetScale / CockroachDB / Tiger Data). Invoke when the user says \"design this schema\", \"r

Deep Lake data architecture specialist for Hivemind - the 7-table ColumnDef schema, `USING deeplake` DDL, FLOAT4[768] embeddings, additive schema healing, append-only version-bump writes, indexing (deeplake_index BM25 / `<#>` vector / hybrid), DeeplakeApi querying, SQL guards, dataset versioning, an

The embedding model selection and runtime specialist. Covers choosing and calling a hosted provider (OpenAI text-embedding-3, Cohere embed-v3/v4, Voyage AI) or running a self-hosted local model (transformers.js), dimension and cost tradeoffs, batching, caching to avoid re-embedding identical text, a

Cognitive-layer specialist for the deploying product, covering coach/agent routing, prompt cascade, RAG/GraphRAG, three-tier memory, observability, evaluation, multimodal pipeline, orchestration, matching, and onboarding as stack-neutral architecture, plus SvelteKit +server.ts streaming LLM response

Neon Postgres + Drizzle ORM specialist for the SvelteKit/Vercel stack migrating off Supabase. Invoke for \"set up Neon\", \"pick a connection driver for Vercel\", \"review this Drizzle schema/migration\", \"is push safe here?\", \"wire pgvector for RAG\", \"do we need RLS without Supabase?\", \"migr

Retrieval specialist for an app - owns Postgres full-text search (tsvector, websearch_to_tsquery, ts_rank), pgvector semantic recall, Reciprocal Rank Fusion hybrid search, optional cross-encoder reranking, chunking strategy, and recall/precision evaluation with golden query sets. Neon Postgres plus

TARS, the session manager for the Claude Code desktop app: turns the owner's asks into chips, manages every session to merged and deployed, fields their questions, makes manager-level decisions, uses CASE for proportion and scoped decisions, and KIPP as the Feature Desk, runs a real 10-minute tick a

Vector and embedding storage specialist - schema and column design, index selection (HNSW vs IVFFlat, distance operators), hybrid lexical+vector search at the storage layer, migrations, and dataset versioning. Neon Postgres plus pgvector plus Drizzle is the primary option for this stack; Deep Lake a

Compare, configure, and troubleshoot AI-assisted development tools and write bounded task briefs for autonomous coding agents.

Provides recommendations and configuration guides for AI stacks, including model selection, gateways, cost optimization, and local LLM workflows.

Provides specialized knowledge of the Bifrost AI gateway architecture, API surfaces, plugin systems, and versioning policies.

Configures Cursor agents, rules, hooks, MCP, and SDK automation for enhanced IDE productivity.

Designs PostgreSQL data layers including schema design, indexing, migrations, performance auditing, and serverless platform selection.

Design and maintain Deep Lake datasets using ColumnDef schemas, additive schema healing, versioned writes, and hybrid indexing.

Helps agents select, configure, and debug embedding models across hosted providers and self-hosted local runtimes.

Reviews, audits, and extends the AI cognitive layer including RAG, prompt cascades, three-tier memory, and observability.

Helps agents manage Neon Postgres and Drizzle ORM within SvelteKit and Vercel environments, including migrations and Supabase transitions.

Designs and audits retrieval systems including full-text search, pgvector semantic recall, hybrid RRF fusion, and reranking strategies.

Build a bounded RAG feature using an accepted corpus, access contract, and measured retrieval and answer evaluation.

Design source-grounded data models, access boundaries, indexes, and migration proposals without executing database changes.

Transforms a Claude Code session into a manager that orchestrates sub-agents, manages session lifecycles, and monitors continuous integration.

Orchestrates Claude Code Desktop sessions by managing parallel work chips, monitoring CI status, and coordinating decision-making via CASE and KIPP.

Design vector storage schemas, indexes, and migrations for pgvector, Deep Lake, Qdrant, and managed vector services.