Agent Setup

View and change Naga settings, add MCP tool services and import skills

Fix web search when a relay or proxy model backend returns empty results

Diagnose and repair repo setup, git hooks and guarded commit or push flows

Write and debug agent hooks that turn repeated rule breaks into hard gates

Route coding agent setup problems to fixes for plugins, statusline and models

Schedule a one-time local ping that opens a new usage window after quota resets

Restructure agent instruction files with progressive disclosure and no lost information

Set up isolated profiles to switch between coding agent model providers safely

Configure a larger, model-aware context window for the Codex CLI and verify it

Audit and switch model providers in OpenClaw config files and diagnose gateway failures

Fix OpenClaw model errors such as 401 invalid tokens and unsupported thinking levels

Set up and customize a coding agent statusline showing directory, model and tokens

Check when Codex usage limits reset and how much quota remains on each account

Switch evo backends and check prerequisites, auth and install setup

Verify phone, Telegram, Slack, Discord or email channels through an outbound check

Run Claude Code in headless, container and CI setups, with auth and git workarounds

Change voice and system settings for your assistant from the chat

Explain what the client apps can and cannot do, such as device location access

Answer questions about setup, config, billing and hosting from live sources

Set up voice in chat, including Talk shortcuts, microphone access and text-to-speech

Run a wizard that builds model routing and capability tier settings for agent teams

List AI CLI tools, their models and subscriptions to choose agent models

Switch an agent's CLI and model live, including turning thinking on or off

Update an installed Aegis method pack to the latest version on this host

Audit and rewrite AGENTS.md so it keeps only lasting principles

Add guardrails so Codex avoids scope creep, stale edits and fake completion

Add Telegram emoji reaction directives to assistant reply text

Apply upstream framework updates to skills and scripts while keeping your files

Repair Codex Desktop on Windows after updates, restoring hidden models and fast mode

Audit CLAUDE.md, AGENTS.md and skills for overconstraint, conflicts and redundancy

Set up the full agent harness for a repo with docs, dev stack, e2e gate and lints

Set up a block policy file that limits which Simulink blocks an agent may use

Import OpenClaw memories, skills, allowlists and workspace assets into a new agent

Schedule a delayed restart of the agent service after config changes

Create an agent harness from a plain-language request and manage its learning

Write AGENTS.md and CLAUDE.md index snippets that route tasks to the right skills

Store your TypeSafe or OpenRouter API key for Jev tools without exposing it

Install, configure and troubleshoot the FlowDown AI app on iOS and macOS

Get an overview of Chorus, its common tools and routing to stage skills

Install and use the chorus CLI to manage agents, env vars and MCP calls

Set permission boundaries, secret protection and review gates for coding agents

Install OpenClaw on a Mac mini and keep the local agent running reliably

Answer install and configuration questions about MateClaw from its own docs

Configure review servers, Python tools and the environment for ARIS research skills

Install, upgrade and remove agent components from the zylos command line

Restart Claude Code cleanly after settings, hook or keybinding changes

Upgrade Claude Code to the latest version with a graceful shutdown and restart

Set up a built-in browser console for chatting with your agent without outside services

Turn on Remote Control for every active Claude Code session in the desktop app