AGENTS.md

AI Agent Instructions — DDoS Protection Platform

This file is read automatically by AI coding agents (OpenAI Codex, Claude, GPT-4o, etc.) that support AGENTS.md at the repository root.

You MUST read and follow project-docs/AI_INSTRUCTIONS.md before making any change to this repository.

Mandatory Reading Order

1. project-docs/AI_INSTRUCTIONS.md — rules, constraints, style
2. project-docs/OVERVIEW.md — architecture & tech stack
3. project-docs/REPORT.md — current implementation status & known issues
4. project-docs/TODO.md — open tasks (check before starting work)
5. project-docs/ROADMAP.md — planned features & architecture direction

Quick Rules

• All new documentation goes in project-docs/ — never create .md files at the repo root.
• Run cd backend && pytest -v before and after any backend change.
• Never use shell=True in subprocess calls; validate IPs with the ipaddress module.
• Every API endpoint must require JWT auth except /api/v1/auth/token and /register.
• Every DB query for ISP-specific data must filter by isp_id.
• Update project-docs/CHANGELOG.md and project-docs/TODO.md with every change.

Documentation Index

All documentation: project-docs/INDEX.md