cursorrules

A set of mandatory instructions for creating security threat models, which identify ways a system could be attacked or misused.

In plain words
What is it for?
Use it when analyzing an application's security risks through a staged threat-modeling process. It first asks for an operating mode and whether each stage should receive a critic review.
Why use it?
It gives the agent a defined process before analysis begins, including choices about user involvement and review. This helps keep threat-modeling work consistent.

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/ensingm2/ai-threat-modeling-rulesets/cursorrules
Clone the repo
git clone --depth 1 https://github.com/ensingm2/AI-threat-modeling-rulesets

Made for: Cursor.

Per session 945 This file is loaded in full into every session.
When invoked 945 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 $0.00945 $0.00945
Opus 5 $0.00473 $0.00473
Sonnet 5 $0.00189 $0.00189
Haiku 4.5 $0.00094 $0.00094

Measured 2d ago against content hash 601c17e27483, 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 2d 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.

.cursorrules · 95 lines

How it starts

The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Threat Modeling Framework

Project: LLM Instruction Sets for Threat Modeling
Creator: Mike Ensing ([email protected])


⚠️ MANDATORY STARTUP SELECTIONS (BEFORE ANY WORK)

CRITICAL: You MUST ask BOTH questions before beginning Stage 1. This is absolute and non-negotiable.

Step 1: Ask BOTH Questions (USE THIS EXACT SCRIPT - MANDATORY)

Copy this prompt VERBATIM when starting a threat model:


Before we begin, I need two selections:

1. Operational Mode:

  • Collaborative Mode: Active user engagement, clarifying questions, stage approvals
  • Automatic Mode: Autonomous operation, reasonable assumptions, no user interaction

2. Critic Review Mode:

  • With Critic Review: Each stage undergoes adversarial quality validation
  • Without Critic Review (Recommended): Stages proceed directly without separate validation phases

Important Note: Critic Review mode will be especially valuable when multi-agent support is added, enabling a separate agent to perform independent validation. For single-agent runs, the benefit of self-critique is generally not worth the increase in runtime and API requests. Recommendation: Without Critic Review unless you have a specific need for it.

Please specify: [Collaborative/Automatic] and [With Critic/Without Critic]


Step 2: Load mode file and confirm

  1. Load modes/collaborative-mode.md OR modes/automatic-mode.md
  2. Confirm BOTH selections before proceeding

Step 3: Begin Stage 1

PROHIBITED:

  • ❌ Starting Stage 1 without asking about operational mode
  • ❌ Starting Stage 1 without asking about critic review mode
  • ❌ Assuming either preference without explicit user selection
  • ❌ Asking about mode but forgetting to ask about critic review
  • ❌ Using your own wording instead of the exact script above
  • ❌ Asking the two questions in separate prompts (MUST be combined)

Quick Reference

  1. Read: .ai-instructions/core/entry-point.md
  2. Read: .ai-instructions/skills/workflow-guide.md
  3. ASK USER BOTH: Operational mode + Critic review mode (see above)
  4. Load: .ai-instructions/modes/[selected-mode].md
  5. Confirm selections
  6. Read ALL user-provided source files (enumerate directories, read every file - no exceptions)
  7. Begin Stage 1

Read the full file on GitHub · 95 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. 2d ago First seen · 95 lines · 945 tokens per session scan A 601c17e27483

Subscribe to this mod's changes

cursorrules is a cursor rule published in the GitHub repository ensingm2/AI-threat-modeling-rulesets (12 stars, last pushed 6mo ago), licensed MIT. It adds 945 tokens to every session, about $0.0047 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.