EmberNest
Get your assurance reportBook a conversation

Standards

Governing coding agents against the OWASP agentic top 10.

The OWASP Top 10 for Agentic Applications, published in December 2025, names the main risks when AI agents act on their own. Coding agents meet several of them directly, because they read code, call tools and change systems.

Last reviewed .

The risks most relevant to coding agents

  • ASI02 Tool misuse
  • ASI03 Identity and privilege abuse
  • ASI04 Agentic supply chain vulnerabilities
  • ASI05 Unexpected code execution
  • ASI06 Memory and context poisoning
  • ASI08 Cascading failures
  • ASI10 Rogue agents

Read the full list and definitions on the OWASP site. These seven are the ones coding agents meet most directly.

Controls that reduce them

  • Work inside an approved scope
  • Isolated worktrees
  • Gates before merge
  • A test that defines done
  • A full record of what each agent did
  • Scoped, time-limited permissions per agent (roadmap)

EmberNest provides the first five today. Per-agent permissions are on the roadmap and we will describe them as live once they ship.

Download the checklist
We email you the file and use your address for nothing else unless you ask. Privacy notice
Get your free assurance report

This page is not legal advice.

All standards

Talk through the standards that apply to you.

Book a conversation