Packages
Browse packages indexed by the registry.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Local-first MCP server giving AI coding agents fast, structured, and semantic context over any codebase.
Macros for axum-router-helper
Async Executor Instrumentation Observability Utility A tracing subscriber layer that outputs Tokio's tracing instrumentation to the terminal in a format conducive to debugging.
Unified agent runtime: wraps the codex harness loop, injects recalled context (ContextStore contract), isolates tools in a sandbox.
On-device / local context store for the agent platform, backed by aria memo (SQLite). Implements aria-agent-core's ContextStore contract; the cloud uses Postgres + pgvector instead.
Pluggable sandbox abstraction for the agent platform. Docker (default), Kata, Cube providers.
High-performance AI inference auditor for ARIA BRC-121
Canonical JSON serialization and SHA-256 hashing for ARIA BRC-121
RFC 6962 Merkle tree for ARIA BRC-121 audit records
Generic adaptive sequencing engine — zero dependencies, domain-agnostic. Suggest(), feedback(). Works from item one.
CLI for the Aria transformer runtime — run Spec-faithful Φ-cycles with invariant checking, trace export, and configurable schedules
Simulated operator backends for the Aria transformer — pluggable optical unitary, JEPA predictor, graph editor, and diffuser implementations
Spec-faithful Aria state machine engine — optical–JEPA–graph dynamical system core with formally verified safety invariants
Python bindings for the Aria transformer runtime — Spec-faithful Φ-cycles with invariant checking
WebAssembly surface for the Aria transformer runtime — run Spec-faithful Φ-cycles in the browser
Aria JSON telemetry — closed operator envelopes, hosted command list, work gateway. 560 catalog identities; one transform.
It's a library AND a binary, but at what cost?
Rust port of ARIA Query.
Lightweight in-process builtin agent for aria-router
Model selection algorithms for aria-router
YAML v0.3 config for aria-router
Shared types and RouterError for aria-router
Boolean decisions and projections for aria-router
Process-out agent extensions (pi, deepseek-harness)
OpenAI data plane and management API for aria-router
Request/response plugins for aria-router
OpenAI-compatible upstream forwarding for aria-router
Semantic routing signals for aria-router
aria2-rust core download engine: multi-protocol segmented downloads, rate limiting, configuration, session persistence, and BitTorrent seeding
Multi-protocol networking stack for aria2-rust: HTTP/HTTPS client, FTP/SFTP, full BitTorrent (DHT/PEX/MSE), and Metalink V3/V4 parser
Remote control server for aria2-rust: aria2-compatible JSON-RPC 2.0, XML-RPC, and WebSocket download management
Yet Another Aria2 JSON-RPC Client.
The Aria2r download tool, written in Rust, is a lightweight and versatile solution that supports multiple protocols (HTTP/HTTPS, FTP, SFTP), as well as BitTorrent and Metalink.
UniFFI FFI surface for the agent platform (Swift / Kotlin bindings).