📦 Package skillz

The skillz package provides intelligent utilities for extracting and normalizing technical skills from raw text (e.g., resumes or job descriptions) using a pluggable LLM (Large Language Model) client. It also estimates proficiency levels for recognized skills based on contextual evidence.


Interfaces

type Processor

type Processor interface {
	ExtractAndNormalize(ctx context.Context, text string) ([]string, error)
	ExtractProficiencies(ctx context.Context, text string, knownSkills []string) (map[string]string, error)
}

Defines a high-level interface for extracting and normalizing skills, as well as estimating proficiencies.


type LLMClient

type LLMClient interface {
	CallLLM(ctx context.Context, prompt string) (string, error)
}

Abstracts interaction with any LLM backend (e.g., Gemini, OpenAI). All implementations must return raw model output from a text prompt.


Structs

type LLMProcessor

type LLMProcessor struct {
	aliasMap  map[string]string
	caser     cases.Caser
	llmClient LLMClient
}

An LLM-based Processor that: