Search

One command, one page: every token your AI coding tools have spent

Decision engine that treats LLMs as next-action predictors with constrained structured output

Local-first memory layer and protocol for AI coding agents. Markdown source of truth, SQLite FTS5 index, zero cloud.

Schema, interfaces, and pipeline engine for reviewpipe. Zero ML dependencies.

Agent-agnostic local computer-use CLI

Browser-based coding assistant for AbstractFramework - connects to AbstractGateway for durable agent sessions

Apple Vision OCR provider for DeepSeek Harness: document structure on macOS 26+, line recognition elsewhere, through a prebuilt Swift sidecar

Prebuilt dsh-vision Apple Vision OCR sidecar for darwin-arm64 — resolved as a file path by @phamkhachoabk/dsh-ocr-apple-vision, never imported

Prebuilt dsh-vision Apple Vision OCR sidecar for darwin-x64 — resolved as a file path by @phamkhachoabk/dsh-ocr-apple-vision, never imported

No description available

Automatic cross-session memory for DeepSeek Harness, backed by a local EverOS service.

Checkpoint inspection and context handoffs for Pi coding agents

browser voice control system powered by gemini live api, easily integrate ai voice assistant with access to tools on your web app, custom function calls, coupled with pre-build tools for dom manipulation

Mona Chat Widget Component Library

DevClub UI animated component primitives

Intent-aware prompt rewriter and execution-contract compiler for Pi

Continuous, verbatim context compaction for LLM agents using TypeSafe's Jev model.

dsh Web 插件：在输入框模型名左侧加一个小星星按钮，一键把当前草稿交给模型改写成结构化提示词，可一键撤回。

Command-line client for Nimbase company memory

n8n community node for Ragas RAG evaluation (LLM-as-judge metrics)