LibbieAI

Deterministic tooling for local LLM agents. Lean, audited, payments-in-USD.

Most agent frameworks burn tokens routing every action through the LLM. We build the opposite: deterministic dispatch with zero LLM calls in the hot path. The core (swiszard router + local semantic memory) is free and MIT. The first-party handler Packs are paid - they ship as audited Python wheels with structured-JSON output. Every Pack comes with a 7-day money-back guarantee.

Open Core MIT - FREE

libbieai-swiszard

The deterministic tool router + local semantic memory server. One MCP tool slot for file ops, shell, web search, and memory. Local SQLite + nomic-embed-text via Ollama. No telemetry.
github.com/cadenKel/libbieai-swiszard
pip install libbieai-swiszard[all]

Packs

Loading...

How it works

  1. Pay via PayPal Checkout.
  2. You get a one-time download link by email + on the success page.
  3. Install the wheel into your Python environment.
  4. If it does not work for you, email [email protected] within 7 days for a full refund.

Why we built this

Ergonomic handles for your LLM. The swiszard core is free and open-source -- the paid packs collapse multi-turn LLM dances (refactor, grep, git, test-fix, deps, log triage) into one deterministic op. You pay for the work your model would have done anyway. Terms & refund policy.