HighlightCurated by Aramai Editorial

How the Open Knowledge Format can improve data sharing

Google introduces the Open Knowledge Format (OKF) to standardize knowledge representation for AI systems.

The Open Knowledge Format is an open specification that formalizes the LLM-wiki pattern into a portable, interoperable format. It represents knowledge as a directory of markdown files with YAML frontmatter and allows for standardized documentation and data sharing across teams and organizations. The OKF aims to solve the problem of fragmented context landscapes by providing a vendor-neutral, agent- and human-friendly standard for representing metadata, context, and curated knowledge.

Based on: How the Open Knowledge Format can improve data sharing | Google Cloud Blog · cloud.google.com

HighlightCurated by Aramai Editorial

Intentional Arrangement

A Substack publication by Jessica Talisman, MLS, on information architecture and semantic engineering.

The resource covers topics such as ontologies, knowledge graphs, AI, and semantic interoperability. It includes essays and articles on various aspects of digital knowledge ecosystems and their organization. The author shares her expertise in information architecture and semantic engineering, with a focus on intentional arrangement and its applications.

Based on: Intentional Arrangement | Jessica Talisman, MLS | Substack · jessicatalisman.substack.com

HighlightCurated by Aramai Editorial

DuckDB RDF Extension

A DuckDB extension to read and write RDF files directly.

This extension allows reading and writing RDF files in DuckDB, supporting various formats such as Turtle, NTriples, NQuads, and TriG. It uses the SERD library for parsing and writing RDF data. The extension also supports WebAssembly (WASM) builds and compression formats like Gzip and Zst.

Based on: GitHub - nonodename/duck_rdf: RDF file extension for DuckDB. Reads and writes supported · github.com

HighlightCurated by Aramai Editorial

Version Control in TuringDB

TuringDB is a graph database with built-in version control and time travel capabilities.

TuringDB combines graph database capabilities with time travel, enabling isolated workspaces, full change tracking, and safe collaboration. It provides Git-like versioning for graph data, allowing users to explore ideas in parallel, test hypotheses, reproduce past states, and audit data evolution. The system includes core concepts such as commits, main branch, changes, and head, which enable users to safely experiment, audit the entire commit history, and roll back accidental or problematic changes.

Based on: Version Control in TuringDB - TuringDB · docs.turingdb.ai

HighlightCurated by Aramai Editorial

CloakBrowser

A stealth Chromium browser that passes every bot detection test.

CloakBrowser is a drop-in replacement for Playwright with source-level fingerprint patches. It claims to pass all bot detection tests, making it suitable for tasks requiring undetectable browsing. The project includes examples and documentation on how to use the library.

Based on: GitHub - CloakHQ/CloakBrowser: Stealth Chromium that passes every bot detection test. Drop-in Playwright replacement with source-level fingerprint patches. 30/30 tests passed. · github.com

HighlightCurated by Aramai Editorial

Looking to the future: SPLASH

A new project for next-generation information modeling in the age of AI.

The author presents SPLASH, a Standard Pattern Library for Advanced Science and Healthcare. It aims to provide semantic scalability across domains using ontology marking and code generation. The model allows for fine-grained data representation, fractal structures, and entity references with thumbnails.

Based on: Looking to the future: SPLASH - Woland's Cat · wolandscat.net

HighlightCurated by Aramai Editorial

GitHub - DataTreehouse/maplib

A high-performance RDF knowledge graph construction library in Python.

maplib is a Rust-based library for constructing and querying knowledge graphs. It supports SHACL validation, SPARQL and Datalog queries, and can read knowledge graphs from various serialization formats. The library allows users to leverage their existing skills with Pandas or Polars to extract and wrangle data before building a knowledge graph.

Based on: GitHub - DataTreehouse/maplib · github.com

HighlightCurated by Aramai Editorial

Canopy - Runtime Specifications

Documentation for Canopy's runtime engine and supported runtimes.

This resource describes the runtime interface and requirements for Canopy, an open-source agent workspace runtime. It outlines the properties of a Canopy-compatible runtime and lists supported runtimes, including Claude Code as the default. The documentation also covers invocation patterns and Workspace settings for configuring runtime behavior.

Based on: canopy/specs/06-runtime.md at main · coralogix/canopy · github.com

HighlightCurated by Aramai Editorial

Agent Capsule - A pattern for building production AI agents

A pattern for building production AI agents as document folders powered by coding-agent as runtimes.

The Agent Capsule pattern proposes using existing coding-agents as runtime engines and defining agents as folders of documents. This approach aims to simplify the development cycle and enable faster iteration and delivery. The pattern includes a template agent folder, workspaces, and an orchestration layer for managing templates and user workspaces.

Based on: Agent Capsule - A pattern for building production AI agents as document folders powered by coding-agent as runtimes · gist.github.com

HighlightCurated by Aramai Editorial

Coralogix Canopy

An open-source agent workspace runtime for building multi-user AI products.

Canopy is a framework for building multi-user AI agent products where .md files define the agent's behavior, memory, and capabilities. It provisions Workspaces from templates, handles triggers, spawns runtime instances, and routes output. The runtime is pluggable, with Claude Code as the default.

Based on: GitHub - coralogix/canopy: Canopy - the open-source agent workspace runtime · github.com

HighlightCurated by Aramai Editorial

Terraphim AI

A local-first knowledge graph search engine built in Rust and WebAssembly.

Terraphim AI is a local-first knowledge graph search engine that indexes, searches, and traverses graphs on-device. It uses Aho-Corasick pattern matching and deterministic graph embeddings to provide fast and explainable search results. The engine ships with nine capabilities, including role-based knowledge graphs, graph embeddings, and multi-haystack search.

Based on: Terraphim AI · terraphim.ai

HighlightCurated by Aramai Editorial

abtop - Like htop, but for AI coding agents

A real-time monitoring tool for AI coding agents, similar to htop.

abtop is a command-line interface (CLI) tool that monitors and displays information about running AI coding agents in real-time. It supports multiple agents, including Claude Code, Codex CLI, and OpenCode, and provides features such as token tracking, context window monitoring, rate limit detection, and more.

Based on: GitHub - graykode/abtop: Like htop, but for AI coding agents. Monitor Claude Code & Codex CLI sessions, tokens, context window, rate limits, and ports in real-time. · github.com