ry-ops

Aiana — AI skill for Claude Code

AI community

AI conversation attendant for Claude Code - monitors and records conversations in real-time via Claude Code API.

How to install Aiana

This entry records only its repository, not the path inside it, so there is no exact command to give. Open ry-ops/aiana and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Aiana does

AI conversation attendant for Claude Code - monitors and records conversations in real-time via Claude Code API.

Alternatives in AI

  • Manifest — Real-time cost observability for OpenClaw agents — track tokens, costs, messages, and model usage 4.1k ★
  • Sentrux — Real-time architectural sensor that helps AI agents close the feedback loop, enabling recursive self-improveme 2.9k ★
  • CodeIsland — Real-time AI coding agent status panel in your MacBook notch — live status, approvals & replies for 13 AI tool 2.3k ★

README

On Monday you tell Claude Code you use uv not pip; aiana stores it. On Friday, in a new session, aiana recalls it and Claude adds the dependency with uv, as you prefer.

Aiana

A memory for Claude Code. Aiana watches your sessions, remembers what happened, and feeds it back into new ones — so every session builds on the last instead of starting cold. 100% local.

9 MCP tools Python 3.10+ Storage backends 100% local MIT


The idea

Without aiana:  session → do work → forget
With aiana:     session → do work → remember → recall → compound

Your preferences persist, your decisions are kept, and Claude Code learns the way *you* work — all on your own machine.

How it works

A watcher and Claude Code hooks capture conversations; an embedder vectorizes them; storage fans out to SQLite+FTS5, Qdrant, Redis and mem0; nine MCP tools and a context injector read it back into Claude Code.

  • Capture — Claude Code's official hooks (and a Docker-friendly file watcher) record each session.
  • Embed — sentence-transformers (all-MiniLM-L6-v2) turns text into vectors.
  • Store, four ways — SQLite + FTS5 for full-text, Qdrant for semantic search, Redis for cache, and mem0 for extracted/deduplicated memory. Each backend is feature-detected; missing ones are skipped.
  • Recall