v0.0.11 — current alpha (tip of main)

v0.0.11 is where liminal-ai main sits today: MIT-licensed, still alpha. Memory and persona can live in user-global storage while the web UI runs multiple chats with transcription and external API/MCP connectors.

Shipped in this slice#

• Storage layout — ~/.liminal/ for cross-project memory, persona, prefs, vault; per-chat folders for sessions and artifacts; AGENT_STORAGE_LAYOUT=legacy preserves old paths
• Web multi-chat — ChatManager, chat list/switch REST APIs, idle bridge eviction, SSE events scoped to the active chat
• Audio — transcribe_audio, upload attachments, Whisper-turbo defaults via OpenRouter, dictation env knobs, web mic UI
• API connectors — api_connect / api_disconnect / api_list (OpenAPI specs → persisted dynamic tools)
• MCP — mcp_attach registers MCP server tools into the live harness
• Memory — memory_promote, memory_neighbors, consolidate_chat; federated rank scoring; exploratory-turn debias (AGENT_MEMORY_DEBIAS)
• Harness — AGENT_PROMPT_CACHE for provider cache breakpoints; intent inference heuristics + research finalize judge
• Docs — studio product pipeline on vireondynamics.com/roadmap; hosted install URL comments

What you can run today#

Clone the repo, set AGENT_API_KEY in .env, build core + tools, then:

npm run build -w packages/core && npm run build -w packages/tools
npm run web   # or: npm run tui

Or use the hosted installer:

curl -fsSL https://vireondynamics.com/install/install.sh | bash

Full operator docs: docs.vireondynamics.com/liminal/.

Roadmap (not shipped)#

Pre-release history#