composing-a-policy-layer

composing-a-policy-layer is a skill for Claude Code, Codex from trustless-ai/agent-sdk. It costs 94 tokens per session (2,344 once invoked), scanned A, original, Apache-2.0.

A skill for adding an external policy engine to a trustless-ai ERC stack. A policy engine decides whether a proposed agent action may proceed, while verification checks whether records or proofs are valid.

In plain words
What is it for?
It is for adding checks such as amount limits, allowlists, rate limits, sanctions screening, or human approval around agent actions.
Why use it?
It helps keep permission decisions separate from identity, record anchoring, and proof verification.

Skill for Claude CodeCodex

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.

agentmods
npx agentmods add skills/trustless-ai/agent-sdk/composing-a-policy-layer
Any agent
npx skills add trustless-ai/agent-sdk --skill composing-a-policy-layer
Clone the repo
git clone --depth 1 https://github.com/trustless-ai/agent-sdk

Made for: Claude Code, Codex.

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 composing-a-policy-layer

README.md
[![agentmods](https://agentmods.dev/badge/skills/trustless-ai/agent-sdk/composing-a-policy-layer.svg)](https://agentmods.dev/skills/trustless-ai/agent-sdk/composing-a-policy-layer)
Your own site
<a href="https://agentmods.dev/skills/trustless-ai/agent-sdk/composing-a-policy-layer"><img src="https://agentmods.dev/badge/skills/trustless-ai/agent-sdk/composing-a-policy-layer.svg" alt="Measured on agentmods" height="20"></a>
Per session 94 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,344 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00094 $0.02344
Opus 5 $0.00047 $0.01172
Sonnet 5 $0.00019 $0.00469
Haiku 4.5 $0.00009 $0.00234

Measured 5d ago against content hash 6d22a5baf513, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

composing-a-policy-layer 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 5d 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/skills/composing-a-policy-layer/SKILL.md · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Composing a policy layer

A policy engine answers one question about a proposed agent action: does it proceed or not. Amount caps, allowlists, rate limits, sanctions screening, a human in the loop gate. This is real and useful work, and it sits next to the ERC stack rather than inside it. This skill is about wiring one in without dissolving the guarantees the stack is built on.

The subject here is any external policy engine. Bastion and CAPV appear once, at the end, as a concrete worked example, not as the point.

The one rule to internalize first

The verifier decides validity. The policy engine decides whether to proceed. These are different questions and they must stay in different components.

A policy layer that also starts deciding what an ERC record means, whether an inference proof is good, or whether an action should be anchored has stopped composing and started wrapping. Wrapping feels tidy because one object now speaks for the whole flow, but it inherits every breaking change in every ERC it speaks for, and it moves the trust boundary back onto itself. The whole reason the stack exists is that a reader does not have to trust any one component. A policy layer that swallows the primitives quietly undoes that.

1. Where a policy layer sits

Beside the primitives, on the path, gating the transition. Not above them, not translating them.

ERC-8004 identity        proposed action        ERC-8263 anchor
    (who)          ->        (what)        ->      (recorded)
                               |
                         Policy Guard
                     (proceed or do not)
                               |
                       ERC-8274 verify
                       (inference proof)

The identity registry still says who the actor is. The verifier still says whether a proof is valid. The anchor still records the timeline. The policy Guard reads what it needs from those and returns a single verdict on whether this action is allowed to continue. It does not re-answer any of their questions.

Read the full file on GitHub · 86 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. 5d ago First seen · 86 lines · 94 tokens per session scan A 6d22a5baf513

Subscribe to this mod's changes

composing-a-policy-layer is a skill published in the GitHub repository trustless-ai/agent-sdk (2 stars, last pushed 3d ago), licensed Apache-2.0. It adds 94 tokens to every session and 2,344 once invoked, about $0.0005 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 2026-08-31.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

local-ai-agents

Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…

microsoft/ai-agents-for-beginners · 200 tokens

next-cache-components-adoption

Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…

vercel/next.js · 95 tokens

next-cache-components-optimizer

Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…

vercel/next.js · 170 tokens

next-partial-prefetching-adoption

Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…

vercel/next.js · 103 tokens

chronicle

Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…

microsoft/vscode · 72 tokens