Atuin v18.13 adds AI shell commands, faster search daemon, and PTY proxy

Search improvements with daemon
The daemon, previously experimental, now maintains a hot, in-memory search index powered by a modified version of nucleo, which implements the same algorithm as fzf. Configurable multipliers for frequency, recency, and frecency scoring are available. Enable with:
search_mode = "daemon-fuzzy"
[daemon]
enabled = true
autostart = trueWith autostart enabled, Atuin manages the daemon lifecycle. The daemon also enables syncing records without shell interaction, keeping remote machines updated.
Atuin AI for shell commands
Press ? on an empty prompt, describe what you want, then press enter to execute or tab to edit. Press f to follow up with adjustments or questions. Enable with:
[ai]
enabled = trueAccuracy comes from frontier models combined with a dataset powered by man pages and command outputs. Safety checks flag dangerous commands (like wiping data or restarting deployments) requiring double enter to execute. Data privacy is minimal by default—only OS and shell info—with optional permissions for directory path, contents, git status, etc.
PTY proxy (hex)
Hex enables Atuin popups to render over previous terminal output without clearing it, unlike previous fullscreen or inline modes. It's a lightweight PTY proxy that behaves like minimal tmux but doesn't affect scrollback or break terminal features. Enable one-off with atuin hex or permanently with:
eval "$(atuin hex init)"Future uses include searching/syncing command outputs, supporting shells without hooks, and readline-esque features.
Other updates
New authentication for hosted sync service (Atuin Hub) adds Google/GitHub login and email recovery. Existing credentials still work. Hub isn't open source yet due to rapid changes. Self-hosting remains unchanged.
📖 Read the full source: HN AI Agents
👀 See Also

Flotilla v0.5.0 Overhauls Background Execution to Beat Claude SDK Credit Caps
Flotilla v0.5.0 replaces sequential agent execution with non-blocking parallel loops, 30-minute per-agent timeouts, and local delegation to cut SDK credit usage.

ThumbGate Implements Tsinghua's Natural-Language Agent Harness Pattern for AI Safety
The open-source tool ThumbGate implements the Natural-Language Agent Harness pattern from Tsinghua's NLAH paper, mapping four components: contracts to prevention rules from thumbs-down feedback, verification gates to PreToolUse hooks, durable state to SQLite+FTS5 lesson database, and adapters to MCP server adapters for multiple AI coding agents.

ClawCut Proxy Released on GitHub to Optimize OpenClaw for Small LLMs
ClawCut is an experimental proxy that manipulates, injects JSON calls, and extracts JSON clutter from OpenClaw to reduce cognitive overload on small models (7B-8B) running on limited hardware.

Revise: AI Editor Built with Agentic Coding Tools and Y.js CRDT
Revise is an AI editor for documents built from scratch over 10 months using agentic coding tools, with a custom word processor engine and rendering layer that only uses Y.js for the CRDT stack. It integrates multiple AI models including GPT-5.4 variants and Claude models for proofreading and revision.