Security Threat Model Skill (OpenAI) for Claude Code: What It Does and How to Install It

Makes Claude write a threat model for your repo or a path in it: trust boundaries, assets, attacker abilities, abuse paths and mitigations.

Installs: 8.5K; on skills.sh
Stars: 28.4K; on GitHub
Files: 4; besides SKILL.md
Updated: Oct 2026; last repo commit

The skill is a curated entry in OpenAI's skills catalog for Codex, written as an Agent Skill. Claude collects the system model from your repo, lists trust boundaries and entry points, ranks a small set of abuse paths by likelihood and impact, and writes <repo-or-dir-name>-threat-model.md. Before the final report it asks you 1 to 3 questions and waits for your answers.

It loads only when you explicitly ask for a threat model. SKILL.md says not to use it for general architecture summaries or code review. The folder holds SKILL.md, 2 reference files, a LICENSE and an agents/openai.yaml, with no scripts. The repo README says the openai/skills repository is deprecated in favor of openai/plugins.

Good for

AppSec engineers who need a first-draft threat model of a codebase
Developers asked to threat model a service before launch
Teams that want assumptions listed and confirmed before a report is final

Install security-threat-model

Claude Code

Copy this prompt.

Install the security-threat-model skill in Claude Code. Run this command:

npx skills add https://github.com/openai/skills --skill security-threat-model

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

Pick one route. The README documents $skill-installer for Codex, which installs curated skills by name. Skills CLI and folder routes install only this skill. The repo README says openai/skills is deprecated in favor of openai/plugins.

Built from the package data we hold. Not yet tested by us.

What's in the skill

When it loads

Repository-grounded threat modeling that enumerates trust boundaries, assets, attacker capabilities, abuse paths, and mitigations, and writes a concise Markdown threat model. Trigger only when the user explicitly asks to threat model a codebase or path, enumerate threats/abuse paths, or perform AppSec threat modeling. Do not trigger for general architecture summaries, code review, or non-security design work.

Sections in SKILL.md

Quick start
Workflow
1) Scope and extract the system model
2) Derive boundaries, assets, and entry points
3) Calibrate assets and attacker capabilities
4) Enumerate threats as abuse paths
5) Prioritize with explicit likelihood and impact reasoning
6) Validate service context and assumptions with the user
7) Recommend mitigations and focus paths
8) Run a quality check before finalizing
Risk prioritization guidance (illustrative, not exhaustive)
References

Key facts

Author: OpenAI (the openai/skills catalog; repo owner)
Repo: openai/skills
License: Apache-2.0 (the skill folder's LICENSE.txt)
Installs on skills.sh: 8,519 (Nov 2, 2026)
GitHub stars: 28,468 for the whole repo (Nov 2, 2026)
Last repo commit: Oct 28, 2026
Audits on skills.sh: Gen Agent Trust Hub: pass, Socket: pass, Snyk: pass
What it includes: 5 files: SKILL.md, LICENSE.txt, agents/openai.yaml and 2 reference files; no scripts
Repo status: The README says the repository is deprecated; use openai/plugins

What it includes

Finds components, data stores, integrations and entrypoints; excludes out-of-scope paths.
Files · 1. Scope and system model
Lists trust boundaries as edges between components, plus assets and entry points.
Read · 2. Boundaries, assets, entry points
Describes realistic attackers and states what they cannot do.
3. Attacker capabilities
Lists a small number of high-quality threats tied to assets and boundaries.
4. Abuse paths
Rates likelihood and impact as low, medium or high and sets an overall priority.
5. Prioritize
Summarizes key assumptions, asks 1 to 3 questions and waits before the final report.
If you cannot answer, the report lists the assumptions that remain.
Try: “Threat model the services/ directory of this repo”
6. Confirm with you
Separates existing mitigations from recommended ones and ties them to locations.
7. Mitigations and focus paths
Checks coverage, then writes <repo-or-dir-name>-threat-model.md.
The report follows an output contract in references/prompt-template.md: 11 headed sections from Executive summary to Notes on use.
Write · 8. Quality check and output
Optional checklist of asset categories and controls.
Read · Reference: security-controls-and-assets.md

From the official docs

Questions

What does the security-threat-model skill do?

It makes Claude write a threat model for a repo or a path in it. The model covers trust boundaries, assets, entry points, abuse paths, risk ratings, mitigations and paths for manual review. It is saved as a Markdown file.

How do I install it?

Use the install box at the top of this page. The OpenAI README documents the skill-installer for Codex. For Claude Code, the box has the skills CLI and folder routes.

When does it trigger?

Only when you explicitly ask to threat model a codebase or path, list threats or abuse paths, or do AppSec threat modeling. The description says not to use it for architecture summaries, code review or non-security design work.

Does it ask me anything?

Yes. SKILL.md says to ask 1 to 3 targeted questions about environment, exposure, auth and data sensitivity, and to pause for your answers before the final report.

Does it work in the Claude app?

Not recommended. The skill reads a repository and writes <repo-or-dir-name>-threat-model.md into it, so use it where Claude has your code on disk, such as Claude Code. The Claude app upload route is not shown on this page.

Does it need an API key?

SKILL.md names no API key and the folder bundles no scripts.

Is it safe?

The folder bundles no scripts. Claude reads your repo to build the model and writes one Markdown report file, so check the report before sharing it, because it describes weaknesses in your system.

Is the openai/skills repo still maintained?

The README says the repository is deprecated and points to openai/plugins for current Codex skill and plugin examples. This skill is still in openai/skills, and the repo was last pushed on Oct 28, 2026. We did not find a security-threat-model folder in openai/plugins, so there is no newer copy to point you to. The install routes on this page use openai/skills.

Information

updated Nov 2, 2026

Repository: openai/skills
Folder: skills/.curated/security-threat-model

openai/skills skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 3

openai/skills README · github.com
security-threat-model SKILL.md · github.com
prompt-template.md · github.com