Ruby gem that analyzes an agent contract (system prompt, tools, policy gates) and generates a structured evaluation pipeline: action plan with SCOPE/INJECT/ROLE/PII/TOOL vectors, executable suite.jsonl with forbidden_tools filled, and remediations for prompt and policy gates. Parallel to security_pentest_planner, applied to LLM agents.

Required Ruby Version

>= 3.0.0

Authors

Saulo Filho

Versions

  1. 0.2.0

    1. source

      1. 0.2.0 source October 21, 2026 (36.5 KB)
  2. 0.1.0

    1. source

      1. 0.1.0 source August 26, 2026 (18 KB)

Pushed by

saulofilho

SHA 256 checksum

← Previous version