Skip to content

patent v0.2.0

Choose a tag to compare

View all tags
@r14dd r14dd released this 03 Jun 12:09
· 61 commits to main since this release
v0.2.0
e34eb89

patent 0.2.0 adds support for any LLM backend and sharpens the verdict.

Highlights

  • Use any LLM, not just Ollama. --api-base (with --api-key or the
    OPENAI_API_KEY env var) points the verdict at any OpenAI-compatible API —
    OpenAI, OpenRouter, Groq, vLLM, LM Studio, llama.cpp. Local Ollama stays the
    default, and --fast still skips the LLM entirely.
  • Better verdicts on close matches. A near-identical result is no longer
    reported as "no matches found," and a single very-close match (similarity
    ≥ 0.70) is treated as a crowded space rather than open.
  • More robust fallback. Any backend failure — unreachable endpoint, wrong
    model or key, malformed response — degrades to a search-only result instead of
    aborting the run.

Install

cargo install patent

Or grab a prebuilt binary / one-line installer from the assets below.

Usage

# local Ollama (default)
patent "kubernetes log viewer"

# any OpenAI-compatible API
patent "kubernetes log viewer" --api-base https://api.openai.com/v1 --model gpt-4o-mini
Assets 2
Loading