HN Showcase
Back to browse
GitHub Repository

Github Copilot skills for threat modeling, scanning, triage, patching, plus an autonomous scanning harness you can /customize

1 starsPython

GitHub Copilot port of Anthropic's AI vulnerability discovery harness

by dreis_sw·Jun 8, 2026·2 points·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidBig Brain

Makes Anthropic's security harness accessible to Copilot users who lack Claude Code access.

Strengths
  • •Real engineering work mapping Claude Code agents to Copilot CLI skills.
  • •Includes sandboxed pipeline with Docker and gVisor for safe autonomous scanning.
  • •Six interactive slash commands for threat modeling, scanning, triage, and patching.
Weaknesses
  • •Explicitly a port, not novel architecture — follows Anthropic's reference design.
  • •Author admits it's a reference that won't work on every codebase out of the box.
Category
Security
Target Audience

Security engineers and developers with GitHub Copilot subscriptions

Similar To

Anthropic defending-code-reference-harness · CodeRabbit · Snyk

Post Description

Last week, Anthropic released https://github.com/anthropics/defending-code-reference-harne..., a reference harness for autonomous vulnerability discovery that uses Claude Code agents to find, verify, and patch memory-safety bugs. I wanted to use it but I only have access to GitHub Copilot.

This is a port of that harness to the GitHub Copilot CLI. PORTING-PLAN.md covers the decisions made to map the handful of features that work differently between Claude Code and the Copilot CLI.

The result is a working reference for anyone who wants to build autonomous security agents on Copilot, tracking Anthropic's approach as closely as possible.

Feedback welcome!

Similar Projects