regaan/Basilisk AI Security Scan

AI/LLM red teaming for CI/CD. Detects prompt injection, guardrail bypasses, and OWASP LLM Top 10 vulns.

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
targetTarget URL or API endpoint to scanyes
providerLLM provider (openai, anthropic, google, azure, ollama, github, custom)noopenai
modelModel name overrideno""
api-keyAPI key for the LLM provider (use GitHub Secrets)no""
modeScan mode: quick, standard, deep, stealth, chaosnoquick
fail-onMinimum severity to fail the pipeline: critical, high, medium, low, infonohigh
outputReport format: html, json, sarif, markdownnosarif
modulesComma-separated list of attack modules to run (default: all)no""
evolveEnable genetic prompt evolutionnofalse
generationsNumber of evolution generationsno3
posture-onlyRun guardrail posture scan only (no attacks)nofalse
baselinePath to baseline SARIF file for regression detectionno""
anthropic-api-keyAnthropic API key (alternative to api-key for multi-provider setups)no""
google-api-keyGoogle API key (alternative to api-key for multi-provider setups)no""
github-tokenGitHub token for GitHub Models (free via github.com/marketplace/models)no""
namedescription
exit-codeScan exit code (0 = pass, 1 = fail)
findings-countTotal number of findings
critical-countNumber of critical findings
high-countNumber of high findings
report-pathPath to the generated report
posture-gradeGuardrail posture grade (A+ through F), only for posture-only mode