Agent Integration Overview
AxonX gives agents discoverable Task contracts, execution interfaces, and research records. Use an existing agent host to operate AxonX, or use built-in sessions in Studio. Both paths share research services and workspaces; different parties manage model configuration and sessions.
Choose a path
| External agent | Built-in agent | |
|---|---|---|
| Entry point | Hosts such as Codex or Claude Code | Studio → Agent, or the agent_chat Job |
| Model configuration | Managed by the external host | Server-side Claude Agent SDK configuration |
| AxonX connection | Direct CLI access or service /mcp | Built-in component invokes configured Jobs through Dispatcher |
| Instructions | AxonX Skill and development guide | Optionally load the development guide bundled with the package |
| Tool scope | Available CLI commands or public service MCP catalog | job_tools, SDK tools, and permission settings |
| Session records | Managed by the external host | Managed by AxonX session storage |
| Reading path | External agents → MCP integration | Built-in configuration → Built-in usage |
External integration does not require CLAUDE_CODE_* settings or built-in guide loading. The built-in agent's eight default Job tools query tasks and files; the public external MCP catalog may also include submission, cancellation, and installation. Actual capabilities depend on the selected path's configuration and schemas.
The shared research loop
- Establish the service address, credentials, and workspace; query machine resources and installed plugins.
- Discover Task definitions and read their actual input and output schemas.
- Reuse successful upstream tasks and execute only the stages the experiment needs.
- Keep the returned
task_idandrun_id, and wait for that execution's terminal state. - Inspect logs, metadata, artifacts, and lineage before forming a research conclusion.
Keep the same --target across stages when using direct CLI access. Studio forwards requests through its same-origin backend to the selected machine. Tasks, data, and plugins live in the execution service's environment; agent integration does not migrate them automatically.
Continue by goal
- Operate a service or develop research plugins: external agents, development and operations guide.
- Analyze existing records in a browser: built-in agent usage.
- Design ablations and independent confirmation: experiment design and confirmation.
- Look up connections, responses, and events: MCP integration, API overview.