OpenClaw Skill Governance (Balanced Dynamic Core Pool)

Manages skill lifecycle and routing using a balanced governance model with status transitions, core pool tuning, and failure-based quarantine.

Before installing or enabling this skill, ensure the governance scripts it references are available and have been audited. The package contains only documentation — it does not include the .ps1 scripts it tells you to run. If you plan to use it, either (a) obtain and review the scripts from a trusted source and store them in the workspace with restricted permissions, or (b) implement your own vetted scripts that follow the described policy. Limit the skill's workspace access (or run it in a sandbox) so the registry and any sensitive files cannot be altered without review. Backup skill-registry.json before running any reconciliation/cleanup actions, and prefer an operator-driven audit step for promotions/demotions rather than fully autonomous execution until you have vetted the implementation. If you cannot verify where the scripts come from, do not enable automatic routing or give the agent permission to fetch or execute remote code to satisfy the missing scripts.

Type: OpenClaw Skill Name: skill-governance-balanced Version: 1.0.1 The provided files consist of metadata and documentation for a skill governance system. The `SKILL.md` describes the skill's purpose, policy, and a list of local PowerShell scripts for managing skill states. It explicitly claims 'No outbound network calls', 'No credential collection', and 'Scripts only read/write local workspace files'. There is no evidence of prompt injection against the AI agent, obfuscation, or any other malicious intent within the provided content. While the described use of dynamic input like `-TaskText` for PowerShell scripts could introduce vulnerabilities if the scripts themselves are poorly implemented, the documentation itself does not contain malicious instructions or code.