> ## Documentation Index > Fetch the complete documentation index at: https://docs.mutagent.io/llms.txt > Use this file to discover all available pages before exploring further. # Supported integrations > Every LLM provider, sandbox provider, trace source, apply target and channel Mutagent works with, and where each one is set up. Check here whether Mutagent works with your stack. Each row links to its setup page. | Group | What it is | Where it is configured | | - | - | - | | [LLM providers](#llm-providers) | The model services Mutagent calls with your keys | Per workspace | | [Sandbox providers](#sandbox-providers) | Where cloud sessions and managed agents run | Shared by every workspace | | [Trace sources](#trace-sources) | Where Helix reads your agent's runs from | Per project, in `.mutagent/config.yaml` | | [Apply targets](#apply-targets) | Where an approved fix is written | Per project, in `.mutagent/config.yaml` | | [Channels](#channels) | GitHub and Slack, to start and follow runs | The app once per organization; its repositories on GitHub | ## LLM providers You bring your own keys. An LLM provider added to a workspace belongs to that workspace only; two workspaces never share one. Add them in the web app under [Configuration › LLM Providers](https://app.mutagent.io/configuration/llm-providers) or with `mutagent providers add`, or copy the keys you signed in with in Helix using `mutagent providers mirror`. Every entry is tested against the provider before it is stored, and a stored entry can be re-tested with `mutagent providers test `. `mutagent providers list --models` shows the models each entry offers. | LLM provider | `--provider` | Fields you supply | Status | | - | - | - | - | | OpenAI | `openai` | API key | Available | | Anthropic | `anthropic` | API key | Available | | Google Gemini | `google` | API key | Available | | Moonshot AI (Kimi) | `moonshot` | API key | Available | | Z.AI (GLM) | `glm` | API key | Available | | Z.AI (Coding Plan) | `zai-coding-plan` | API key from a GLM Coding Plan subscription | Available | | DeepSeek | `deepseek` | API key | Available | | xAI (Grok) | `xai` | API key | Available | | Azure (OpenAI or Anthropic models) | `azure` | API key, resource endpoint, deployment name, API version (optional) | Available | | Google Vertex (Gemini or Anthropic models) | `vertex` | Service-account credential, project ID, region | Available | | Amazon Bedrock (Anthropic models) | `bedrock` | Bearer API key, region, inference profile (optional) | Available | | Custom (OpenAI- or Anthropic-compatible endpoint) | `custom` | Base URL, API key, API format | Available | | OpenRouter (models from many vendors) | `openrouter` | API key, model allow-list (optional) | Available | OpenAI, Anthropic, Google Gemini, Moonshot AI, Z.AI (GLM), DeepSeek and xAI also take an optional base URL override and a model allow-list. For Z.AI (GLM), the default model is `glm-5.3`. Z.AI (Coding Plan) offers the GLM models your plan includes. [LLM provider setup](/platform/providers/setup) lists every field. Helix on your machine does not need any of these: it calls the provider directly with the key in your environment or the one you set with `/login`. Workspace LLM providers are used by the hosted parts, such as cloud sessions and managed agents. ## Sandbox providers A sandbox provider runs cloud sessions and managed agents: an isolated machine with Helix and your agent's tools on it. | Sandbox provider | Scope | Status | | - | - | - | | Mutagent Cloud (`mutagent-cloud`) | The platform's own sandbox provider, and the default. One provider, shared by every workspace. There is nothing to set up per workspace. | Early access | Cloud sandboxes, and the GitHub and Slack runs that use them, are not open to every account yet. They open to every account after production acceptance. LLM providers and Environments are per workspace because they hold your keys and your secrets. The sandbox provider holds neither, so one list serves every workspace; the keys and Environment a session needs are passed in from the workspace that launches it. `mutagent sandbox providers` lists the sandbox providers, and `mutagent sandbox presets` lists the presets: named sandbox definitions a cloud session can be launched from. See [Sandbox providers](/platform/sandbox/providers). ## Trace sources Helix reads traces where they already are. It does not copy them to Mutagent. | Trace source | What is read | How to connect | Status | | - | - | - | - | | Claude Code sessions | Transcripts under `~/.claude/projects/` | `platform: claude-code`, no credentials | Available | | Codex sessions | Transcripts under `~/.codex/sessions/` | `platform: codex`, no credentials | Available | | Helix sessions | Helix's own session files under `~/.mutagent/sessions` | Read by Helix directly, no credentials | Available | | JSONL / NDJSON files | One run per line, gzip supported | `platform: local-jsonl` with `paths` | Available | | [Langfuse](/integrations/sources/otel) | Traces in a Langfuse project | `platform: langfuse`, endpoint, project, key by env var name | Available | | [OpenObserve](/integrations/sources/otel) | Traces and OTel log-event streams | `platform: otel`, endpoint, headers by env var name | Available | | [SigNoz](/integrations/sources/otel) | Traces over the OTel query API | `platform: otel`, endpoint, headers by env var name | Available | | [Dash0](/integrations/sources/dash0) | Traces in a Dash0 dataset | See the Dash0 page | Coming soon | Local sources are described in [Local trace sources](/integrations/sources/local). Langfuse and OpenObserve can also be connected to a workspace once, with `mutagent integrations sources add`, so Mutagent pulls the traces for you instead of Helix reading them from your machine. ## Apply targets After you approve a fix, Helix writes it to a target. Every target lands the change as a GitHub pull request on your repository; nothing is written to your default branch. With no target configured, every stage is report-only. | Kind | Targets | Setup | | - | - | - | | Agents defined as markdown | Claude Code, Codex, Cursor, OpenCode, and skill directories. Pi and Hermes are coming soon. | [Markdown agent targets](/integrations/targets/harness) | | Agents defined as code | Mastra, Claude Agent SDK. DeepAgents and Pydantic AI are coming soon. | [Code targets](/integrations/targets/code) | | Report only | Propose changes, never write | `platform: report-only` | ## Channels | Channel | What it does | Status | | - | - | - | | [GitHub](/integrations/github) | Connect the Mutagent GitHub App to your repositories so an issue, a mention or a failed CI run starts a Helix run. Runs started from an issue post their result back as a comment | Early access | | [Slack](/integrations/slack) | Connect the Mutagent Slack app so a mention starts a Helix run and the reply comes back in the thread | Early access | Connect them in the web app under **Configuration › Integrations** (each has a **Connect** button), or with `mutagent gateway connect github` and `mutagent gateway connect slack`. Each app is installed once for your organization. The repositories Mutagent can use are the ones you give the Mutagent GitHub App access to on GitHub. GitHub and Slack connect through the gateway, which starts each run in a cloud sandbox. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.