Pinned Loading

1. llm-wiki llm-wiki Public

   LLM-powered Wikipedia knowledge agent — MCP integration, Ollama/OpenAI support, semantic search, streaming answers with cited sources

   Python 1

2. Reduction Reduction Public

   Five-layer token-optimization pipeline for AI agents: optimize every input and output (shell filter, LLMLingua, semantic cache, native prompt caching, Caveman+TOON)

   Python 1

3. Strive_Engineering Strive_Engineering Public

   A loop-engineering runtime: design, verify, and safely run autonomous coding-agent loops (L0→L3). Tested Rust safety core + Python orchestration. Works with free NVIDIA NIM models.

   Python 1

4. SEC_Guardrails_Agent SEC_Guardrails_Agent Public

   Defensive 7-layer runtime guardrails gateway for LLM agents: a reverse proxy that blocks prompt injection (incl. indirect/XPIA), tool misuse, and data exfiltration on every turn. Polyglot — Python …

   Python 1

5. semantic-entropy-gate semantic-entropy-gate Public

   Detect LLM confabulation with semantic entropy (Farquhar et al., Nature 2024) and gate agent actions on it. Zero-dependency Python: API + middleware + CLI, with a full audit trail for every decision.

   Python 1

6. Trajectory_Causal_Attribution Trajectory_Causal_Attribution Public

   Record AI agent trajectories and find which step caused a failure via counterfactual step-ablation replay (P(fail|kept) - P(fail|ablated), Point-of-Commitment, Shapley attribution). Zero-dep Python…

   Python 1