Agenticracy

SKILL.md

---
name: agenticracy-skill
description: Public placeholder skill describing a non-proprietary Agenticracy-style audit workflow for grounding claims in evidence, memory, and transparent schema.
metadata:
  author: https://opencollective.com/agenticracy
---
# Agenticracy Public Placeholder Skill

This is a **public placeholder skill** for publishing and handle capture.

## What it is for

Use this skill when you want Agend.MD to produce a **transparent, non-proprietary audit record** that separates:

- memory
- evidence
- inference
- uncertainty
- refusal conditions

It is intended as a safe public draft, not as the private production method.

## Core workflow

1. Identify the subject being audited.
2. Gather only evidence that is explicitly available.
3. Separate facts, claims, and inferences.
4. State uncertainty clearly.
5. Refuse if the task would require secrets, private coefficients, or sensitive profiling.
6. Return a concise audit record with a plain-language verdict.


## Why it exists

- to reserve a public handle
- to document the intended idea safely