Skip to content
AI OS
Claude Code agent mark
Local agent route
Local agentLive executionMission Control

Claude Code execution cockpit

Focused local execution for Claude Code inside AI OS. Workspace, shell, git, process, and file actions run through token-gated local bridge routes while Mission Control stays available on the same page.

Mode
loading
Execution
checking
Missions
idle
Writes
checking
Detecting the local bridge for Claude Code.
Loading Claude Code status

Checking the local bridge before enabling Claude Code Mission Control.

Claude Code execution

LoadingActions gated
Checking Claude Code execution access

Checking Claude Code local execution access.

Execution results

Results show bounded local output, safe path labels, warnings, and recovery steps for the last Claude Code action.

Action controls are disabled because local Claude Code execution is gated.
local Claude Code execution is gated

Claude Code Mission Control

Idle
Writes gated
Claude Code Mission Control waiting for bridge

Mission reads start after the local Hermes bridge is available.

Browser output stays bounded to route status, mission labels, and sanitized bridge state. Raw credentials, tokens, local paths, and prompt payloads are not surfaced.
Claude Code execution uses token-gated local bridge routes for workspace, shell, git, process, and file actions.
Mission Control remains available alongside execution so planning, focused goals, and local action results stay in one Claude Code route.