All integrations¶
This index lists the integration names documented in this section. Installation commands, authentication requirements, diagnostics, and operational boundaries are documented on the linked category page.
| Integration | Function | Category |
|---|---|---|
| Agent Client Protocol, ACP | Connect external coding agents and expose normalized sessions | Protocols and tools |
| Agent2Agent, A2A | Discover, call, and serve interoperable agents | Protocols and tools |
| Amazon Bedrock AgentCore | Experimental remote code-interpreter sandbox | Sandboxes |
| Anthropic Claude Agent SDK | Run the Claude Agent SDK through a SuperQode runtime adapter | Coding agents |
| Apple Container | Experimental macOS-native container status integration | Sandboxes |
| Apple Foundation Models | Route supported utility work to an on-device Apple model | Models and inference |
| Arize Phoenix | Export harness traces through an OTEL collector path | Observability |
| AutoResearch | Edit and evaluate optimization candidates through Claude Code | Optimization |
| Claude Code over ACP | Connect an authenticated Claude Code agent | Coding agents |
| CocoIndex Code | Add AST-aware semantic code search | Protocols and tools |
| Codex over ACP | Connect the Codex CLI as an ACP coding agent | Coding agents |
| Cognee | Use a separately installed local or cloud memory provider | Memory |
| Daytona | Execute commands in a remote development sandbox | Sandboxes |
| DeepAgents | Execute a HarnessSpec through the DeepAgents runtime | Runtimes and harnesses |
| Discord | Control the channel daemon through a Discord bot | Remote interfaces |
| Docker | Execute commands in a local container sandbox | Sandboxes |
| DwarfStar | Connect a compatible local inference server | Models and inference |
| E2B | Execute commands in an E2B cloud sandbox | Sandboxes |
| Exa | Add the optional Exa search client | Models and inference |
| GEPA | Optimize text artifacts through reflective search | Optimization |
| GEPA meta-harness | Maintain and evaluate an optimization candidate frontier | Optimization |
| GEPA Omni | Explore with GEPA, AutoResearch, and GEPA meta-harness | Optimization |
| GitHub Copilot SDK | Connect models and sessions from a Copilot account | Coding agents |
| Google Agent Development Kit | Execute a HarnessSpec through Google ADK | Runtimes and harnesses |
| Google Antigravity CLI | Connect the signed-in agy coding agent | Coding agents |
| Google Antigravity SDK | Execute the Antigravity SDK through a runtime adapter | Coding agents |
| Harness Protocol | Operate native and external harnesses through one lifecycle | Protocols and tools |
| Hugging Face Hub | Search and download hosted model artifacts | Models and inference |
| Hugging Face Tau | Connect the Tau harness through a read-only integration | Runtimes and harnesses |
| Kimi Code | Connect the Kimi CLI over ACP | Coding agents |
| LangSmith observability | Export normalized runs as LangSmith run trees | Observability |
| LangSmith sandbox | Execute commands in an experimental LangSmith sandbox | Sandboxes |
| LiteLLM | Route model requests across supported providers | Models and inference |
| LM Studio | Connect a local OpenAI-compatible model server | Models and inference |
| Local memory | Store explicit memories in the local SuperQode data directory | Memory |
| Local OS sandbox | Apply macOS Seatbelt or Linux Bubblewrap isolation | Sandboxes |
| Logfire | Mirror run summaries and events as spans and logs | Observability |
| Mem0 | Add the optional Mem0 memory provider | Memory |
| MetaHarness | Export and optimize a HarnessSpec through Superagentic MetaHarness | Optimization |
| Microsoft Model Context Protocol, MCP | Connect tool servers and expose harness tools | Protocols and tools |
| MLflow | Export run artifacts and metrics | Observability |
| MLX and MLX-VLM | Run supported models locally on Apple Silicon | Models and inference |
| Modal | Execute commands in a Modal cloud sandbox | Sandboxes |
| Monty Python REPL | Add bounded Python calculation to a harness | Protocols and tools |
| Ollama | Connect locally served models | Models and inference |
| Omnigent | Import compatible Omnigent agent specifications | Runtimes and harnesses |
| OpenAI Agents SDK | Execute a HarnessSpec through OpenAI Agents | Runtimes and harnesses |
| OpenAI Codex SDK | Connect the authenticated Codex SDK runtime | Coding agents |
| OpenResponses | Use the built-in OpenResponses-compatible gateway | Models and inference |
| OpenTelemetry | Export live harness spans to an OTLP collector | Observability |
| Plugins | Add repository-owned tools, commands, and hooks | Protocols and tools |
| Podman | Execute commands in a local container sandbox | Sandboxes |
| PydanticAI | Execute a HarnessSpec through PydanticAI | Runtimes and harnesses |
| Qwen Code | Connect the first-party Qwen Code ACP server | Coding agents |
| RLM Code | Execute recursive workflows through the RLM Code backend | Runtimes and harnesses |
| Runloop | Execute commands in an experimental Runloop devbox | Sandboxes |
| SGLang | Connect a local SGLang inference endpoint | Models and inference |
| SkillOpt | Coordinate optimization and review of markdown skills | Optimization |
| Slack | Control the channel daemon through Slack Socket Mode | Remote interfaces |
| SpecMem | Store project-scoped HarnessSpec memory | Memory |
| Supermemory | Add the optional Supermemory provider | Memory |
| Telegram | Control the channel daemon through a Telegram bot | Remote interfaces |
| Text Generation Inference, TGI | Connect a local or remote TGI endpoint | Models and inference |
| Transformers | Run supported Hugging Face models through a local Python stack | Models and inference |
| vLLM | Connect a local or remote vLLM endpoint | Models and inference |
| Vercel Sandbox | Execute commands through the Vercel Sandbox CLI | Sandboxes |
| Web TUI | Host the terminal interface in a browser | Remote interfaces |
| xAI Grok Build | Connect the authenticated Grok coding agent | Coding agents |
Provider-specific model routes are maintained in the provider catalog. ACP agent packages are maintained in the ACP catalog. Those dynamic catalogs are not expanded into separate left-navigation entries.