Official Python SDK -- connect any AI agent to Crucible security infrastructure in 2 lines.
Status: Under active development. The SDK is being built alongside the Crucible core framework. Star this repo to get notified when the first release drops.
from crucible_sdk import CrucibleGuard
guard = CrucibleGuard(api_key="sk-...")
result = guard.scan("https://my-agent.com/api/chat")
print(result.grade)Two lines. Your agent gets a full security scan with a letter grade and actionable report.
- Drop-in middleware for LangChain, CrewAI, and OpenAI Agents
-
- Continuous monitoring with scheduled scans and alerting
-
- Runtime guard that intercepts and blocks malicious inputs