Pinned Loading

1. agentshield-benchmark agentshield-benchmark Public

   Open benchmark for AI agent security tools — prompt injection, data exfiltration, tool abuse, provenance

   TypeScript 27 10

2. qllm-sim qllm-sim Public

   Python 1

3. jevc jevc Public

   Compile agent policy prose into deterministic verdict programs: narrow evidence questions for the model, the verdict computed in code. Install: npm i -g jev-compiler

   TypeScript 8

4. gitmemory gitmemory Public

   Git-versioned, contiguity-checked memory for AI agents: verbatim transcripts, byte-offset recall, no LLM

   Python 19