Ctrlk
For the complete documentation index, see llms.txt. This page is also available as Markdown.

Overview

Comprehensive probe library for agentic AI Red Teaming

Akto Probe Library is the AI Red Teaming framework for teams protecting AI agents, LLM components, and MCP endpoints. Write custom probes to discover vulnerabilities specific to autonomous systems including prompt injections, tool abuse, and permission escalation.

Probe Library consists of three blocks:

  1. Probe Library with 1000+ built-in probes for agentic security

  2. YAML probe editor for custom probe creation

  3. Sample agent component requests and responses for scanning

PreviousConceptsNextProbe YAML

Last updated