Anthropic Claude provider for TS-DSPy - type-safe LLM interactions with structured outputs, streaming, and tool calling

Google Gemini provider for TS-DSPy - type-safe LLM interactions with structured outputs, streaming, and tool calling

OpenAI and OpenAI-compatible (Ollama, Groq, vLLM, OpenRouter) provider for TS-DSPy - type-safe LLM interactions with structured outputs, streaming, and tool calling

Effect-native DSPy — programming, not prompting, language models

Core library for building type-safe LLM applications with structured input/output signatures, runtime validation, and reasoning patterns in TypeScript

AI-guided LLM optimization. Install → Tell Claude 'Read .claude/agents/iris.md' → Claude becomes your optimization guide. DSPy prompts, Ax hyperparameters, local LLMs, federated learning. You talk, Iris handles the rest.

DSPy-style automated prompt tuning. Search over candidate prompt mutations using eval-runner as the fitness function.

AI agent orchestration framework for TypeScript/Node.js - 32 adapters (LangChain, AutoGen, CrewAI, OpenAI Assistants, OpenAI Responses, LlamaIndex, Semantic Kernel, Haystack, DSPy, Agno, MCP, OpenClaw, A2A, Codex, MiniMax, NemoClaw, APS, Copilot, LangGrap

Production-ready examples for @ruvector/agentic-synth - DSPy training, multi-model benchmarking, and advanced synthetic data generation patterns

Node ↔ Python bridge for DSPy. Spawns a uv-pinned Python sub-process that hosts compiled DSPy programs and answers JSON-line RPC. Substrate pick #1 of the AI tech modernization proposal §6.

Effect-native DSPy — programming, not prompting, language models

A 1:1 TypeScript alternative to DSPy — Declarative Self-improving Language Programs

DSPy/ax-style typed signatures and a ReAct agent loop for the AI SDK

Composable, agentic pipelines for MCP — fact-checking (DSPy), RAG (hybrid search), document intelligence (DeepPipe), and clinical voice (Groq/ElevenLabs). 31 AI tools, one command, zero manual setup.

GKOS-Engine — canonical deterministic GKX engine with an optional, proposal-only DSPy intelligence sidecar; offline validation, projection, assessment, graph, and export remain authoritative.

The Dak Stack deepen generator — the self-improving loop's ORIGINATOR (the eval gate is the validator; this proposes what the gate judges). proposeRevision is a GEPA/DSPy-style hill-climb over a skill's natural-language scaffolding (the SKILL.md body): a

Hal9: Create and Share Generative Apps

DSPy.ts 2.0 - Advanced declarative AI framework with multi-agent orchestration, self-learning capabilities, and state-of-the-art optimizers. Powered by AgentDB, ReasoningBank, and Swarm architecture.

SDK for LLM pipelines with Prime Agent, OpenCode, OMP, Codex, Claude Code, Pi, and Zod

Axle core runtime for building agentic LLM applications.