Agent guides
Server running? One command wires your agent to it.
$ rapid-mlx serve qwen3.6-35b-4bit # in another terminal, pick your agent: $ rapid-mlx agents claude-code --setup $ rapid-mlx agents codex --setup $ rapid-mlx agents pi --setup
env block in ~/.claude/settings.json, with a diff preview and a backup.~/.codex/config.toml.~/.pi/agent/models.json, with a diff preview and a backup.
--setup asks the running server which model it serves (and,
for configs that record it, the model's context window). Pass
--model or --base-url (default
http://localhost:8000/v1) to override, and
--dry-run to preview without writing.
rapid-mlx start <agent> starts the server and
configures the agent in one step.
Supported agents
| Agent | Setup | Writes |
|---|---|---|
| Claude Code | rapid-mlx agents claude-code --setup or rapid-mlx launch claude-code | ~/.claude/settings.json (env block) |
| Codex CLI | rapid-mlx agents codex --setup | ~/.codex/config.toml + rapid-mlx-model-catalog.json (or under $CODEX_HOME) |
| Pi | rapid-mlx agents pi --setup | ~/.pi/agent/models.json (or $PI_CODING_AGENT_DIR) |
| OpenCode | rapid-mlx agents opencode --setup | ~/.config/opencode/opencode.json |
| Hermes Agent | rapid-mlx agents hermes --setup | ~/.hermes/config.yaml (or $HERMES_HOME) |
| DeepSeek Harness | rapid-mlx agents dsh --setup | ~/.dsh/cordis.patch.yml + .credentials.yaml (or $DSH_HOME) |
| Qwen Code | rapid-mlx agents qwen-code --setup | ~/.qwen/settings.json |
| Kilo Code | rapid-mlx agents kilo-code --setup | ~/.config/kilo/kilo.json |
| Aider | rapid-mlx agents aider --setup | prints OPENAI_API_BASE, OPENAI_API_KEY, AIDER_MODEL |
| OpenHands | rapid-mlx agents openhands --setup | prints LLM_BASE_URL, LLM_API_KEY, LLM_MODEL |
| Continue.dev | rapid-mlx agents continue --setup or rapid-mlx launch continue-dev | ~/.continue/config.json (a rapid-mlx model entry) |
| Cursor | rapid-mlx launch cursor --server-url https://… | Cursor's settings. Needs a public HTTPS URL. |
| Cline | rapid-mlx launch cline | Cline's settings |
Framework profiles (rapid-mlx agents langchain,
pydanticai, smolagents) print environment
variables (OPENAI_BASE_URL …) instead of writing files. See the framework
guides. Other OpenAI-compatible clients (GitHub Copilot, the OpenAI
SDK with gpt-oss) have hand-written guides in the sidebar.
How setup treats your existing config
Each agent's setup uses one of three flows:
Preview + backup
Claude Code · Pi · DeepSeek Harness · Continue
- Prints an exact diff, then asks
Apply this configuration? [y/N]. Use--yesin scripts; without it, a non-interactive run writes nothing. - Copies the old file to
<file>.bak.<unix-time>. - Writes atomically, and refuses if the file changed after the preview.
- Checks
/healthand/v1/modelsafterwards (--no-checkskips this).
Merge
Codex CLI · OpenCode · Hermes · Qwen Code · Kilo Code
Merges the rapid-mlx keys into the existing file and keeps your other keys. A list the template sets is replaced, not merged. For Qwen Code that's modelProviders.openai, so other OpenAI-provider entries there are dropped. Hermes's cli toolset list is the exception, and is merged. TOML/YAML comments are not kept, and no backup is made, so copy the file first if it matters. --dry-run says what would change and writes nothing.
Print env vars
Aider · OpenHands
Prints export lines for your shell and writes no files.
Test an integration
$ rapid-mlx agents opencode --test
--test runs the agent's integration checks against the
running server. Where the profile has a headless mode, that includes a
real one-shot task. Pass --agent-version when an agent's
config format depends on its version.