Releases·Tool Release
Simon Willison releases llm-keys-ui 0.1, a plugin for adding API keys to machines via browser instead of pasting into agent sessions
Simon Willison released llm-keys-ui 0.1, a plugin for his LLM CLI that serves a web UI for storing API keys on a machine without pasting them into agent sessions. It targets users running Codex Remote coding agents controlled from a phone, letting the agent fetch keys via llm keys get commands.
0 engagement1 sourceSun, Sep 20, 2026, 07:22 PM
AI Overview
- Simon Willison released llm-keys-ui 0.1, an LLM plugin providing a web interface for saving API keys without pasting them into ChatGPT-driven agent sessions
- Running 'uvx --with llm-keys-ui llm keys-ui --all' serves the key-entry interface over local network or Tailscale device IPs while controlling machines remotely with Codex
- Agents later retrieve stored keys via shell commands like 'llm keys get anthropic'
Related
Products·launched
GetKeys — CLI and agent tool to create, rotate, copy, and revoke API keys without opening dashboards
GetKeys is a local app with a CLI and agent integration for managing API keys (create, rotate, copy, revoke) so developers and agent builders avoid navigating service dashboards. It targets the workflow where users otherwise paste API keys into chats or fetch them manually for their agents. Data is stored encrypted locally on the user's own disk.
GetKeys is a local app with a CLI and agent integration for managing API keys (create, rotate, copy, revoke) so developers and agent builders avoid navigating service dashboards. It targets the workflow where users otherwise paste API keys into chats or fetch them manually for their agents. Data is stored encrypted locally on the user's own disk.
2 engagement1 sourceredditSun, Oct 11, 2026, 08:39 AM
Products·launched
web-edit: web UI for viewing and editing files in a folder, intended for Tailscale use
web-edit is a small Python web app providing a browser UI for browsing and editing files in a folder, built to run over Tailscale so the author can edit files from his phone while working with Codex Remote. It was itself built by Codex and GPT-6 Astra Extra High. The problem it solves: editing files (like credentials) without passing their contents through the LLM.
web-edit is a small Python web app providing a browser UI for browsing and editing files in a folder, built to run over Tailscale so the author can edit files from his phone while working with Codex Remote. It was itself built by Codex and GPT-6 Astra Extra High. The problem it solves: editing files (like credentials) without passing their contents through the LLM.
0 engagement1 sourcerssMon, Oct 12, 2026, 03:54 PM
Products·launched
Lupin: Run Claude Code on any LLM via proxy
Lupin is a proxy that lets you use Claude Code with models like GPT-5.6 Sol, Kimi K3, DeepSeek Flash, or local models without changing your Claude Code setup. It forwards requests to providers that speak the Anthropic API, preserving your MCPs, skills, and other configurations.
Lupin is a proxy that lets you use Claude Code with models like GPT-5.6 Sol, Kimi K3, DeepSeek Flash, or local models without changing your Claude Code setup. It forwards requests to providers that speak the Anthropic API, preserving your MCPs, skills, and other configurations.
13 engagement1 sourceredditSun, Aug 9, 2026, 10:18 AM
News·Incident
llama.cpp warns of RCE risk when --tools or -ag run without API key
A llama.cpp security warning highlights that running the server with --tools or -ag flags without an API key exposes it to remote command execution. Users are urged to configure agents and setups securely.
A llama.cpp security warning highlights that running the server with --tools or -ag flags without an API key exposes it to remote command execution. Users are urged to configure agents and setups securely.
8 engagement1 sourceredditWed, Aug 5, 2026, 01:51 PM
Releases·Tool Release
Developer releases Local Ollama Chat VS Code extension for local LLM coding assistance
A developer launched 'Local Ollama Chat', a VS Code extension connecting the editor's agent chat to a local Ollama server, citing lag and JSON-garbled output in the existing Ollama extension. It reads workspace files and proposes file edits as reviewable diffs, keeping code entirely local. The project is open source and the author is soliciting feedback.
A developer launched 'Local Ollama Chat', a VS Code extension connecting the editor's agent chat to a local Ollama server, citing lag and JSON-garbled output in the existing Ollama extension. It reads workspace files and proposes file edits as reviewable diffs, keeping code entirely local. The project is open source and the author is soliciting feedback.
3 engagement2 sourcesredditFri, Sep 11, 2026, 08:24 AM