injectshield
io.github.bch1212/injectshield
95
security score
Prompt-injection firewall for AI agents — scan untrusted text before LLM calls.
Security scan
Last scanned 2 hours ago — static analysis of the published manifest.
- lowManifest makes network requests.