mirror of
https://github.com/openclaw/openclaw.git
synced 2026-10-03 01:29:56 +00:00
* feat(plugins): show declared capabilities in plugin details * docs(plugins): add concise setup guides for empty detail pages * chore(workboard): refresh generated manifest formatting * fix(ui): inline customization settings and prioritize capabilities * fix(plugins): keep invalid display metadata nonfatal * fix(ui): retain plugin context in inline alerts * chore(workboard): refresh generated settings bundle reference * fix(ui): load customization controls from plugin settings * chore(workboard): regenerate UI bundle with pinned dependencies * test(plugins): complete staged overview metadata fixture * fix(tooling): include plugin capabilities in extracted PR wrapper * test(gateway): include selected plugin capabilities in discovery fixture * fix(plugins): keep remote details available with invalid UI metadata |
||
|---|---|---|
| .. | ||
| assets | ||
| api.ts | ||
| image-generation-provider.test.ts | ||
| image-generation-provider.ts | ||
| index.test.ts | ||
| index.ts | ||
| onboard.ts | ||
| openclaw.plugin.json | ||
| package.json | ||
| provider-catalog.ts | ||
| README.md | ||
| tsconfig.json | ||
LiteLLM
Connect OpenClaw to an existing LiteLLM proxy for model routing through a shared endpoint. The plugin supports chat models and image generation when the proxy exposes the corresponding OpenAI-compatible routes.
Get started
Start or obtain access to a LiteLLM proxy, then run:
openclaw onboard --auth-choice litellm-api-key
Configure models.providers.litellm.baseUrl for your proxy and supply its key
through onboarding or LITELLM_API_KEY. Refresh the proxy's catalog with
openclaw models list --provider litellm --refresh.
The proxy owns upstream credentials and model availability. Installing this plugin does not start a proxy.
See the LiteLLM guide for endpoint, image generation, and routing setup.