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.
npx agentmods add instructions/urule-ai/urule/claude-mdgit clone --depth 1 https://github.com/urule-ai/uruleWrote 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.
[](https://agentmods.dev/instructions/urule-ai/urule/claude-md)<a href="https://agentmods.dev/instructions/urule-ai/urule/claude-md"><img src="https://agentmods.dev/badge/instructions/urule-ai/urule/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.02996 | $0.02996 |
| Opus 5 | $0.01498 | $0.01498 |
| Sonnet 5 | $0.00599 | $0.00599 |
| Haiku 4.5 | $0.00300 | $0.00300 |
Grade A, and why
urule CLAUDE.md 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 290 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Urule — AI Assistant Guide
Urule is the open-source coordination layer for AI agents. It composes open-source projects (Keycloak, NATS, OpenFGA, OPA, Temporal, PostgreSQL) into a cohesive platform and only builds the coordination glue. This file helps AI coding assistants understand the codebase and contribute effectively.
Architecture Principles
- Contract-first — Types and interfaces live in library packages (
packages/). Services implement them. - Schema-per-service — Each service owns its PostgreSQL schema. Never query another service's database.
- Event-driven — Mutations publish events to NATS. Consumers are idempotent.
- Services are glue — If a service exceeds ~2000 lines, it's doing too much.
- Auth flow — User → Keycloak JWT → service validates via
@urule/auth-middleware→ governance/authz checks.
Tech Stack
- Language: TypeScript ESM (
"type": "module", strict mode) - HTTP: Fastify 5
- ORM: Drizzle ORM (PostgreSQL 16)
- Events: NATS + JetStream
- Tests: Vitest (unit), Playwright (E2E)
- IDs: ULID via
ulidpackage - Validation: Zod (request bodies)
- Auth:
@urule/auth-middleware(Keycloak JWKS) - Frontend: Next.js 14, React 18, Tailwind CSS, Zustand
Project Layout
packages/ — shared libraries (imported by services)
spec/ — entity types, manifest schema, validators
events/ — NATS event envelope, topics, AuditLogger
authz/ — OpenFGA SDK wrapper
auth-middleware/ — Fastify JWT plugin (Keycloak JWKS)
services/ — backend microservices
registry/ — source of truth (orgs, workspaces, agents, runtimes, providers)
packages/ — package install/upgrade/remove lifecycle
packagehub/ — package discovery, search, metadata
governance/ — OPA policy + OpenFGA authz gateway
state/ — room presence, task ownership, widget state (NATS KV)
plugins/
backstage/ — Backstage catalog sync
apps/
office-ui/ — Next.js 14 frontend (Tailwind, Zustand, React Query)
infra/
compose/ — Docker Compose files, SQL seeds, init scripts
scripts/ — dev-setup, run-phase scripts
e2e/ — backend E2E tests (phase1.test.mjs)
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.
- 4d ago First seen · 290 lines · 2,996 tokens per session scan A 19c078ca00d1
urule CLAUDE.md is an instructions file published in the GitHub repository urule-ai/urule (2 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 2,996 tokens to every session, about $0.0150 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.
Other instructions, from other repositories
neurolink CLAUDE.md
Claude Code instructions for juspay/neurolink, covering claude.md, project overview, critical rules, architecture and pattern: factory + registry.
agent-pulse AGENTS.md
Instructions for barretlee/agent-pulse, covering agents.md, 项目使命, 协作约定, 情报模型与主线 and 事实、判断与热点.
taOS AGENTS.md
AGENTS.md instructions for jaylfc/taOS, covering agents.md and changelog fragments.
taOS CLAUDE.md
Claude Code instructions for jaylfc/taOS, covering claude.md and changelog fragments.
powernode-platform CLAUDE.md
Claude Code instructions for nodealchemy/powernode-platform, covering claude.md, project overview, core models, where guidance lives and memory & knowledge.
almanac CLAUDE.md
Claude Code instructions for corcoran/almanac, covering almanac — agent notes, specs and plans are local-only, code navigation (lsp), prose: comments, changelog, docs and voice for user-facing prose.