auditor

Validate skill outputs against a constitution and task plan.

Updated Feb 3, 2026
One-click install
npx skills add https://github.com/dhruvinrsoni/agentskills-garden --skill auditor-dhruvinrsoni
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: auditor
Source: https://github.com/dhruvinrsoni/agentskills-garden/tree/main/skills/00-foundation/auditor
Command: npx skills add https://github.com/dhruvinrsoni/agentskills-garden --skill auditor-dhruvinrsoni

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill acts as a crucial validation layer, ensuring that all outputs generated by other skills strictly adhere to the project's core Constitution and the predefined task plan, preventing unintended deviations and maintaining integrity.

Core Features & Use Cases

  • Constitutional Compliance: Validates outputs against the principles of Truth, Safety, and Non-Destruction.
  • Plan-vs-Output Alignment: Verifies that executed changes directly match the planned steps, flagging any discrepancies.
  • Protected Terms Enforcement: Prevents unauthorized modification of critical domain-specific or public identifiers.
  • Safety Gate Validation: Confirms explicit user approval for destructive operations before execution.
  • Use Case: After a code generation skill produces a new feature, the Auditor checks if the code adheres to the project's coding standards (Truth), if it only includes necessary changes (Safety), and if any potentially risky operations were approved (Non-Destruction).

Quick Start

Validate the output of the previous operation against the project's constitution and plan.

Frequently Asked Questions about auditor

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

FAQPage Schema
How do I validate code changes against a project constitution?

An auditing skill validates code changes against a project constitution by checking outputs against predefined truth, safety, and non-destruction principles, ensuring strict adherence to your core standards.

How do I verify plan-vs-diff alignment for executed tasks?

Verifying plan-vs-diff alignment requires an auditing tool that retrieves a task plan from a scratchpad and compares executed changes against planned steps, automatically flagging discrepancies for review.

How can I enforce protected terms and prevent unauthorized identifier modifications?

Enforcing protected terms is achieved by configuring a validation layer that prevents the unauthorized modification of critical domain-specific or public identifiers during automated skill execution.

How do I validate safety gates for destructive operations?

Validating safety gates for destructive operations involves using an auditing step that confirms explicit user approval was granted before any potentially risky or destructive changes are executed.

Does constitutional compliance auditing require a scratchpad for plan retrieval?

Yes, constitutional compliance auditing requires a scratchpad for plan retrieval, and it optionally accepts a glossary and user approvals to facilitate comprehensive output validation.

What is the best way to add a validation layer for generated code outputs?

The best way to add a validation layer for generated code outputs is to run an auditing process that checks for constitutional compliance, plan alignment, and explicit safety gate approvals.