Reference
These guides are the same implementation doctrine used by the CLI, MCP clients, and in-app agents. They explain how to build the Compass, Responsibility Graph, Seats, Charters, Stewards, Cascade, Signal, Friction, and Sync rhythm.
Install the CLI and MCPThe staged operating-system setup path used by humans, CLI sessions, MCP clients, and in-app agents.
Field-by-field guidance for authoring a strong Compass document directly, with the full JSON contract and worked examples.
Install the public implementation Skill through its signed, verified GitHub release so a general-purpose agent can run the ROST implementation method.
The evidence hierarchy and provisional-assumption discipline that grounds every company-specific claim made during ROST implementation.
How the AI Chief of Staff chat works in the authenticated app shell and what it can safely do today.
How to build a functions-first graph with seats, owners, Stewards, vacancies, and clean authority.
How to draft strong Charters that work for human, agent, and hybrid seats.
The full Charter document contract field by field — decision authority, escalation, measurables, budget, permissions — with worked examples for human, agent, and hybrid seats.
How to decide whether a seat should be human, agent, or hybrid, and how to go live safely.
The visual journey for adding an agent seat: where to start, choosing a mode, placing the seat, naming a Steward, setup, the safety gates, and go-live.
How to build a custom agent from operational questions through the Charter Builder, tools, dry run, and go-live without writing prompts.
The full agent setup sequence on the CLI/MCP path — seat, steward, job, boundaries, tools, credentials, model, schedule, dry-run, go-live — with the structured model config and access tiers.
The strict versioned JSON format for importing an agent as a draft Seat/setup without credentials, placement ids, or go-live side effects.
How ROST agents operate inside seats, use Charters, report work, and escalate beyond authority.
How to give agents access to tools without exposing raw credentials or expanding authority by accident.
How to think about tool categories available to seats and what each category should be used for.
Install the public CLI, register remote token-backed MCP clients, and find the full command and tool catalog.
Create, import, review, publish, and assign reusable Skills without confusing procedures with authority.
How CLI and MCP agents should propose reusable Skills safely.
CLI and MCP recipes for discovering, reading, dependency-checking, assigning, and locally syncing Skills.
Where CLI sessions, MCP clients, and in-app agents should retrieve ROST guidance before recommending setup changes.
How to connect company goals to seat-level work without turning ROST into a project-management tool.
How to define and read measurables so the company runs on evidence instead of status theater.
How to capture issues with evidence, rank them, and resolve them without losing ownership.
How Stewards review escalations, approve agent boundaries, and keep agents accountable.
How ROST packages company-wide access, Stripe billing, and governed-agent usage without per-human-seat pricing.
How ROST routes authority-changing work through human confirmation, and why agents never approve their own requests.
How to read ROST's command capability classification: the four confirmation classes, the Trusted-eligibility classes, the five-command auto band, and the always-human floor.
How to use Settings as the control plane for company access, channels, providers, tokens, and operating defaults.
How to invite humans into the company without confusing user access with seat accountability.
How ROST should notify humans about decisions, escalations, stale work, and agent boundaries.
How local agent sessions and runner surfaces should operate through ROST without bypassing Charters or audit.
How to use stock agent templates as starting points while preserving seat-specific Charters and human approval.
How users and agents should diagnose common setup, tool, Signal, Friction, and MCP problems.
How ROST cloud agents, local MCP sessions, runners, BYOK, and connected tools handle model-bound data.
How ROST protects tenant data, credentials, tool calls, and human decisions across web, MCP, CLI, and agents.
How Signal, Friction, Cascade, and Sync Briefs turn weekly meetings into decision time.
How ROST should handle companies arriving from another operating-system framework without overusing protected third-party marks.