stride-gpt AGENTS.md

stride-gpt AGENTS.md is an instructions file for Codex, OpenCode from mrwadams/stride-gpt. It costs 4,122 tokens per session, scanned A, original, MIT.

A project guide for AI coding agents and people working on STRIDE-GPT, a tool that creates threat reports for software systems. It explains the repository, its command-line tool, web interface, and shared code.

In plain words
What is it for?
Use it to orient an agent before making changes, understand where the CLI, web interface, and shared library live, and reuse the project's progressive-disclosure approach.
Why use it?
It removes the need to discover the project structure and working rules from the code alone. It also explains how the agent explores code and how to reveal guidance progressively.

Instructions file for CodexOpenCode

Written for Codex and OpenCode: the file is AGENTS.md. Also seen: mentions Claude Code; mentions AGENTS.md.

About the project

STRIDE GPT is an AI threat-modeling tool that analyzes an application and generates threat models and attack trees using the STRIDE security methodology. Security practitioners use it through a command-line interface, interactive terminal, or web interface to assess applications, codebases, generative-AI systems, and agentic-AI systems.

mrwadams/stride-gpt · 1,112 stars · on GitHub · stridegpt.streamlit.app

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 instructions/mrwadams/stride-gpt/agents-md
Clone the repo
git clone --depth 1 https://github.com/mrwadams/stride-gpt

Made for: Codex, OpenCode.

Wrote 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.

agentmods badge for stride-gpt AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/mrwadams/stride-gpt/agents-md.svg)](https://agentmods.dev/instructions/mrwadams/stride-gpt/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/mrwadams/stride-gpt/agents-md"><img src="https://agentmods.dev/badge/instructions/mrwadams/stride-gpt/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 4,122 This file is loaded in full into every session.
When invoked 4,122 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 5.1 $0.04122 $0.04122
Opus 5 $0.02061 $0.02061
Sonnet 5 $0.00824 $0.00824
Haiku 4.5 $0.00412 $0.00412

Measured 6d ago against content hash a1acc50344c1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

stride-gpt AGENTS.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 6d 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.

AGENTS.md · 187 lines

How it starts

The opening of the file, as written. The whole thing — 187 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Orientation for agents (and humans) working in this codebase. Skim the whole file before starting; jump to "Progressive disclosure pattern" if you're looking to reuse that specific idea.

What this project is

STRIDE-GPT is an AI-powered threat modelling tool that produces STRIDE reports for software systems. It ships as two separate products with a shared library between them:

  • CLI (stride_gpt/) — the primary product. Includes an agentic analysis engine that explores a real codebase via filesystem tools and produces a per-subsystem STRIDE report.
  • Streamlit web UI (apps/web/) — a hosted single-shot threat modeller. Takes a text description (or a GitHub URL) and produces a one-shot report. Does not use the agentic engine, by design — filesystem/agentic features don't fit Streamlit Community Cloud-style deployments.

Both surfaces share stride_gpt/core/ for prompts, schemas, and the LLM abstraction.

Repository layout

stride_gpt/                 # CLI package + shared library
├── cli.py                  # Typer commands + interactive REPL
├── prompt.py               # prompt_toolkit completer (slash commands)
├── config.py               # ~/.stride-gpt/config.json + provider registry
├── models.py               # model catalogue (id, default tokens, thinking support)
├── agent/                  # agentic loop (CLI-only by design)
│   ├── loop.py             # run_analysis, per-subsystem agent loop
│   ├── planner.py          # phase 1 — classify app type, propose subsystems
│   ├── context.py          # context window management + compression
│   ├── tools.py            # filesystem tools (read_file, grep, ..., list_references, load_reference)
│   ├── progress.py         # Rich-based progress callbacks
│   └── report.py           # markdown / JSON / SARIF rendering, save/load
└── core/                   # shared between CLI and web
    ├── llm.py              # unified call_llm / call_llm_with_tools via litellm
    ├── schemas.py          # LLMConfig, AnalysisPlan, AnalysisReport, etc.
    ├── prompts/
    │   ├── builder.py      # legacy single-shot prompt builder (web UI uses this)
    │   ├── variants.py     # base_system_prompt(), list_references(), load_reference(), coerce_app_type()
    │   └── threat_model/   # packaged reference cards (see "Progressive disclosure")
    │       ├── base.md
    │       ├── quick_base.md
    │       ├── genai.md
    │       ├── agentic.md
    │       ├── insider_threat.md
    │       ├── mitre_enterprise.md
    │       └── mitre_atlas.md
    └── ...                 # attack_tree, dread, mitigations, test_cases, threat_model

apps/web/                   # Streamlit UI (separate product)
└── ...                     # imports from stride_gpt.core, NOT stride_gpt.agent

tests/                      # pytest, fixtures in conftest.py

Read the full file on GitHub · 187 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. 6d ago First seen · 187 lines · 4,122 tokens per session scan A a1acc50344c1

Subscribe to this mod's changes

stride-gpt AGENTS.md is an instructions file published in the GitHub repository mrwadams/stride-gpt (1,112 stars, last pushed 3d ago), licensed MIT. It adds 4,122 tokens to every session, about $0.0206 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.