cursorrules

A set of rules for the IVD framework, a development system that links an intended change to a written intent file and its checks. It requires the agent to read that file before changing related code.

In plain words
What is it for?
Use it when adding a feature, fixing a bug, or changing code that has an intent.yaml file, so the agent can inspect the requirements and follow the framework's verification loop.
Why use it?
It prevents implementation from drifting away from the stated requirements and verification steps. This makes the intended behavior and constraints available before work begins.

Cursor rule for Cursor

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 rules/leocelis/ivd/cursorrules
Clone the repo
git clone --depth 1 https://github.com/leocelis/ivd

Made for: Cursor.

Per session 5,133 This file is loaded in full into every session.
When invoked 5,133 The same file — it is already loaded in full.
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 $0.05133 $0.05133
Opus 5 $0.02567 $0.02567
Sonnet 5 $0.01027 $0.01027
Haiku 4.5 $0.00513 $0.00513

Measured yesterday against content hash 4f8111ad1a67, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

cursorrules 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 yesterday.

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.

.cursorrules · 426 lines

How it starts

The opening of the file, as written. The whole thing — 426 lines — stays where its author put it; the contents beside it link to each section on GitHub.

IVD Framework — Cursor Rules

Development rules for working on the IVD framework itself.

Both blocks below also serve as the canonical example of what

users should install in their own projects when adopting IVD.

==============================================================================

Source: leocelis/ivd/recipes/agent-rules-ivd.yaml v1.11

Reference: ivd_system_intent.yaml, framework.md, recipes/agent-rules-ivd.yaml

Do not modify by hand — reload from recipe to refresh.

==============================================================================

IVD: Intent-Verified Development Rules

==============================================================================

These rules enforce the IVD verification loop. Do not remove them.

Reference: ivd_system_intent.yaml, framework.md (Principle 4, Principle 6)

IVD Rule 1: Intent Before Implementation

Before implementing any feature, fix, or change that has an associated intent artifact (*_intent.yaml):

  1. READ the intent artifact first

  2. UNDERSTAND the constraints and their test paths

  3. IMPLEMENT using constraint-segmented approach:

    FOR INTENTS WITH 3+ CONSTRAINTS (threshold based on IFScale 2025 degradation data and Liu et al. 2023 lost-in-the-middle effect) — MANDATORY, not optional:

    a. GROUP constraints into functional segments after the stress test. Example groups: input validation | core logic / output | performance | error handling | security b. IMPLEMENT one segment ONLY — do NOT single-pass the full intent c. Before implementing any constraint with conflict_prone: true, inject its anti_pattern text into the active implementation context d. RE-READ the intent constraints for that segment from disk (not from memory — forces attention reset; counters lost-in-the-middle) e. VERIFY the segment satisfies its constraints before proceeding f. Repeat steps b–e for each segment

    Single-pass implementation against 3+ constraints violates Rule 1.

Read the full file on GitHub · 426 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. yesterday First seen · 426 lines · 5,133 tokens per session scan A 4f8111ad1a67

Subscribe to this mod's changes

cursorrules is a cursor rule published in the GitHub repository leocelis/ivd (0 stars, last pushed 1mo ago), licensed MIT. It adds 5,133 tokens to every session, about $0.0257 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.