grim-prompter
Threat modeling for AI coding agents: point Claude Code, Codex or any agent at your repo and get an evidence-traced threat model. 36 architectures, 29 methods,
What is grim-prompter?
Threat modeling for AI coding agents: point Claude Code, Codex or any agent at your repo and get an evidence-traced threat model. 36 architectures, 29 methods, MIT.
Overview of grim-prompter
What it is
Grim Prompter is a threat modeling tool for AI coding agents. It provides structured prompts that agents like Claude Code or Codex can follow to analyze a repository and produce an evidence-traced threat model. The project includes 36 architecture workflows and 29 method and analysis prompts.
Key features
- 36 architecture workflows covering applications, cloud, data/AI, devices, identity, and distributed systems
- 29 method prompts for threat discovery, attackers/abuse, privacy, safety, AI systems, and enrichment/scoring
- Quick and deep analysis modes; deep runs cover six stages: scope, system modeling, threat analysis, challenge, treatment planning, and reporting
- Change review: compare a branch against an existing JSON baseline
- Results saved as SYSTEM_UNDER_REVIEW.md, report.md, and JSON artifacts
- Read-only inspection with file-and-line citations; builds, tests, and network access require explicit authorization
Use cases
Threat model a project using Claude Code, Codex, or another coding agent. Choose a quick model for a specific architecture, a deep six-stage analysis, or a single method (e.g., STRIDE per element). Review how a branch changes an existing model. Use the evidence guide to gather a system-under-review bundle without running an analysis.
Getting started
bashgit clone https://github.com/jmanico/grim-prompter.git ~/grim-prompter# Claude Codemkdir -p ~/.claude/skills && ln -s ~/grim-prompter/skills/threat-model ~/.claude/skills/threat-model# Codex CLImkdir -p ~/.agents/skills && ln -s ~/grim-prompter/skills/threat-model ~/.agents/skills/threat-modelStart a new agent session after installation. For other agents, add the project snippet to your project's AGENTS.md.
License and pricing
Licensed under MIT. Open source.
Based on the project's public README and repository metadata.