BoundaryML/baml
BAML (Basically a Made-up Language) is an AI framework that facilitates reliable AI workflows and agents by transforming prompt engineering into schema engineering for structured output generation.
Awesome Infra for AI › Prompt Management
Jia-Ethan/claude-keysmith
claude-keysmith is a deployment tool for the custom instruction files that drive Claude Code. Rather than editing CLAUDE.md by hand and hoping the result is what was intended, it takes a Markdown source, stores it in a keysmith directory, and inserts a recognizable, removable import block into CLAUDE.md or CLAUDE.local.md. Every operation defaults to a preview: the tool shows exactly what it would write, and only applies the change when explicitly confirmed, which keeps a botched instruction rollout from silently changing how an agent behaves. Installations are scoped — project and local scope affect a single repository, while user scope affects every new session that loads the global instruction file — and a runtime mode additionally aligns the systemPrompt entry in settings.json and installs a managed shell wrapper. Alongside install it offers status inspection and clean revocation, so an instruction set can be removed as deliberately as it was added. It is part of a family of Keysmith tools that do the same job for other agent runtimes, including Codex, Grok Build and ZCode, each targeting that tool's own instruction location. Distribution deliberately avoids piping a script from the network into an interpreter: the documented path is a shallow clone of a stable tag followed by running the Python entry point, with an unsigned desktop build available in beta for macOS and Windows for people who prefer a graphical flow. Written in Python with no dependency beyond the standard library on 3.8 and later, and MIT licensed, it targets teams that treat agent instructions as configuration to be deployed, audited and rolled back rather than as files someone pastes in.
https://github.com/Jia-Ethan/claude-keysmith
Also in Prompt Management
genkit-ai/genkit
Genkit is an open-source framework by Google for building and operating AI-powered applications across multiple languages, featuring unified APIs for various models, structured outputs, multi-modal...
pezzolabs/pezzo
Pezzo is an open-source, cloud-native LLMOps platform for streamlined prompt design, version management, instant delivery, collaboration, troubleshooting, and observability of AI operations.
dot-agent/nextpy
Nextpy is a framework for building self-modifying software, focusing on guardrails, structured outputs, a powerful prompt engine for pre-compiling and session state, and optimized code generation f...
neuron-core/neuron-ai
Neuron is a PHP framework for building and orchestrating AI agents, supporting LLM integration, prompt management, RAG, multi-agent workflows, and observability.
microsoft/aici
AICI (Artificial Intelligence Controller Interface) allows building Wasm-based controllers to constrain and direct LLM output in real time, enabling advanced generation strategies.
patterns-ai-core/langchainrb
Langchain.rb provides a Ruby interface for building LLM-powered applications, offering a unified API for various LLM providers, prompt management, and RAG capabilities.
Open-Source-Legal/OpenContracts
OpenContracts is an open-source document intelligence platform that processes unstructured documents into a programmable citation graph using AI agents, structured extraction, and a Model Context P...