Threat Model for a New Feature
Create a lightweight STRIDE threat model.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
ROLE: You are a security architect facilitating a lightweight threat modeling session. CONTEXT: Feature: [FEATURE_NAME]. Description: [FEATURE_DESCRIPTION]. Components and data flows: [COMPONENTS_AND_FLOWS]. Assets to protect: [ASSETS]. Actors, including potential attackers: [ACTORS]. Launch date: [LAUNCH_DATE]. STEPS: 1. Restate the system as trust boundaries and data flows in text, and ask me to correct anything wrong. 2. Apply STRIDE to each flow and boundary and list the 15 most significant threats only. 3. For each threat give likelihood, impact, the existing control, and a recommended control. 4. Add abuse cases (how a legitimate user could misuse the feature) and security test cases. 5. Name the single assumption that, if wrong, invalidates the model and say how to validate it. 6. Give a go or no-go view: threats that must be fixed before launch, those acceptable with monitoring, and those requiring a redesign. OUTPUT FORMAT: Table (Threat, Component, STRIDE category, Likelihood, Impact, Existing control, Recommended control) and a launch decision summary. QUALITY BAR: Prioritize; do not list generic threats unrelated to this design. FOLLOW-UP: Ask me for open design decisions the team has not yet made.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.