moses-postures

Enforce posture-based governance controls on agent actions and transactions.

1|Updated Mar 12, 2026
One-click install
npx skills add https://github.com/SunrisesIllNeverSee/moses-claw-gov --skill moses-postures
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: moses-postures
Source: https://github.com/SunrisesIllNeverSee/moses-claw-gov/tree/main/skills/moses-postures
Command: npx skills add https://github.com/SunrisesIllNeverSee/moses-claw-gov --skill moses-postures

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Posture controls enforce policy-driven throttling of agent actions, ensuring that transactions and executions align with defined governance rules.

Core Features & Use Cases

  • SCOUT — read-only mode to gather information without state changes.
  • DEFENSE — protects assets by requiring explicit confirmation for outbound transfers and double-checks for large changes.
  • OFFENSE — enables execution within governance constraints with full logging and rationale.
  • Dependency awareness: reads ~/.openclaw/governance/state.json and relies on the moses-governance bundle for state initialization.

Quick Start

Set posture to defense to enable guarded execution and require explicit confirmation for outbound transfers.

Frequently Asked Questions about moses-postures

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I enforce governance controls on agent transactions and runtime execution?

Governance postures control agent actions by enforcing policy-driven throttling on transactions and executions. They ensure runtime orchestration aligns with defined rules through SCOUT, DEFENSE, and OFFENSE levels for state changes and external calls.

What is the difference between SCOUT, DEFENSE, and OFFENSE posture levels?

SCOUT is a read-only mode for gathering information without state changes, DEFENSE requires explicit confirmation for outbound transfers and double-checks large changes, and OFFENSE enables full execution within governance constraints with logging and rationale.

How do I set up governance posture to require explicit confirmation for outbound transfers?

Set the governance posture to DEFENSE to enable guarded execution and require explicit confirmation for outbound transfers. This posture protects assets by double-checking large state changes and external calls before they proceed.

Do I need the moses-governance bundle to initialize posture state?

Yes, you need the moses-governance bundle to initialize posture state. The system reads the governance state from the ~/.openclaw/governance/state.json file and relies on this bundle for proper dependency management and state initialization.

Can I use governance postures to run agents in a read-only mode?

Yes, you can run agents in read-only mode by setting the posture to SCOUT. This allows agents to gather information during runtime orchestration without executing state changes or external calls.

What limitations should I expect when applying posture-based governance to agent actions?

A key limitation is the hard dependency on the moses-governance bundle and a readable ~/.openclaw/governance/state.json file. Without proper state initialization, policy-driven throttling cannot be enforced on agent actions and transactions.