README
Start with the core architecture, CLI model, quick start, commands, policy model, and configuration layout.
Architecture, setup, integrations, and operational guidance for brokered agent access.
OpenScope is easiest to understand when the docs are grouped by intent: start here, understand the architecture, extend it safely, and validate it in practice.
Start with the core architecture, CLI model, quick start, commands, policy model, and configuration layout.
Understand how OpenScope fits into the local OpenClaw workflow and why brokered actions are safer than raw local automation access.
Learn the client-only sandbox model, where the broker stays on the host and the sandbox uses a narrow client surface.
These docs explain the conceptual model behind OpenScope: capability brokering, key containment, bypass resistance, and how OpenScope fits alongside gateways and other security controls.
Why AI gateways are not enough for high-risk workflows and where capability brokering becomes necessary.
Compact visual explanations for architecture difference, bypass risk, execution containment, and key containment.
How OpenScope fits into the wider secure-reach plus brokered-action model.
OpenScope can broker both local app actions and external system actions through the same trust model.
Keep the Jira token in a broker-owned HTTP profile while exposing narrow actions like get issue and search issues.
Use named targets, scoped services, and explicit policy for SSH-backed operations.
Define new app actions in YAML with action-level parameters and outputs.
OpenScope has operational concerns that matter for real deployment: signed runtime packaging, broker startup, validation, and pilot readiness.
Signed macOS packaging, LaunchAgent layout, and installer shape.
Shared validation flows across packaged installs, OpenClaw, NemoClaw, and SSH paths.
Operational guidance for early rollout and validation.
A simple sequence for new visitors.