write-policy

write-policy is a command for Claude Code from Owl-Listener/ai-design-skills. It costs 13 tokens per session (547 once invoked), scanned A, original, MIT.

An AI behavior policy is a written set of rules for how an AI should act, speak, handle uncertainty, and respond to sensitive situations.

In plain words
What is it for?
Use it to define what an AI must do, must not do, needs human approval for, and how it should communicate.
Why use it?
It turns broad safety, ethical, legal, and brand expectations into rules that can be reviewed and tested.

Command for Claude Code

Written for Claude Code: argument-hint in frontmatter.

Part of the ai-alignment-reasoning plugin — 8 skills, 3 commands shipped together

Good fit Use it to define what an AI must do, must not do, needs human approval for, and how it should communicate.

Compare 6 commands from other repositories ↓
Install with agentmods
npx agentmods add commands/owl-listener/ai-design-skills/write-policy
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/Owl-Listener/ai-design-skills

Made for: Claude Code.

Or install ai-alignment-reasoning, the plugin that ships this one along with the rest of its 8 skills, 3 commands.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for write-policy

README.md
[![agentmods](https://agentmods.dev/badge/commands/owl-listener/ai-design-skills/write-policy/github.svg)](https://agentmods.dev/commands/owl-listener/ai-design-skills/write-policy)
Your own site
<a href="https://agentmods.dev/commands/owl-listener/ai-design-skills/write-policy"><img src="https://agentmods.dev/badge/commands/owl-listener/ai-design-skills/write-policy/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for write-policy

Your own site · 80×15
<a href="https://agentmods.dev/commands/owl-listener/ai-design-skills/write-policy"><img src="https://agentmods.dev/badge/commands/owl-listener/ai-design-skills/write-policy.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 13 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 547 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00013 $0.00547
Opus 5 $0.00006 $0.00273
Sonnet 5 $0.00003 $0.00109
Haiku 4.5 $0.00001 $0.00055

Measured 10d ago against content hash 674f96fbc456, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

write-policy scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

claude-plugin/ai-alignment-reasoning/commands/write-policy.md · 60 lines

How it starts

The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are drafting an AI behavior policy. Use only skills from the ai-alignment-reasoning plugin. Follow this process:

Step 1: Establish the Value Foundation

Using value-specification:

  • What are the organisation's core values relevant to AI?
  • How do these translate to AI behavior principles?
  • Create a value hierarchy with clear conflict resolution
  • Identify key stakeholder perspectives (users, legal, brand, ethics)

Step 2: Define Behavioral Boundaries

Using guardrail-design:

  • What will the AI always do? (mandatory behaviors)
  • What will the AI never do? (prohibited behaviors)
  • What requires human approval? (escalation behaviors)
  • What varies by context? (conditional behaviors)
  • Write each rule clearly enough to be implemented and tested

Step 3: Specify Tone and Voice

Using value-specification and guardrail-design (tone guardrails):

  • How should the AI sound? Define voice attributes.
  • What tone shifts are appropriate in different contexts?
  • What language is off-limits?
  • How does the AI handle sensitive topics?

Step 4: Define Transparency Requirements

Using transparency-patterns:

  • What must the AI disclose to users?
  • How should uncertainty be communicated?
  • What source attribution is required?
  • When must the AI identify itself as AI?

Using consent-and-agency:

  • What data does the AI use and how?
  • What consent is required from users?
  • What opt-out mechanisms must exist?
  • What override capabilities must users have?

Step 6: Plan for Harm Prevention

Using harm-anticipation and escalation-design:

  • What harm scenarios has the policy been designed to prevent?
  • What escalation procedures exist?
  • How are incidents reported and handled?
  • What review cadence keeps the policy current?

Step 7: Address Bias

Using bias-detection-design:

  • What bias monitoring is required?
  • How often are bias audits conducted?
  • What mitigation processes exist?
  • Who is accountable for bias-related issues?

Read the full file on GitHub · 60 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 10d ago First seen · 60 lines · 13 tokens per session scan A 674f96fbc456

Subscribe to this mod's changes

write-policy is a command published in the GitHub repository Owl-Listener/ai-design-skills (172 stars, last pushed 3mo ago), licensed MIT. It adds 13 tokens to every session and 547 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.