DEEP-CONTENT

Skill Advisor — Development skill for Claude Code

Development community

Auto-route every Claude Code prompt to the right skill, subagent, or slash command — silently, locally, with zero API keys.

How to install Skill Advisor

This entry records only its repository, not the path inside it, so there is no exact command to give. Open DEEP-CONTENT/skill-advisor and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Skill Advisor does

Auto-route every Claude Code prompt to the right skill, subagent, or slash command — silently, locally, with zero API keys.

Alternatives in Development

Ccusage — CLI for analyzing Claude Code/Codex usage from local JSONL files 11.8k ★
Om Dump — Freeform capture mode 4.6k ★
Sanitize — Detect and redact PII from text files — 15 categories (SSNs, credit cards, API keys, etc.), zero dependencies 3.1k ★

README

skill-advisor

**Auto-route every Claude Code prompt to the right skill, subagent, or slash command — silently, locally, with zero API keys.**

[![CI](https://github.com/deep-content/skill-advisor/actions/workflows/ci.yml/badge.svg)](https://github.com/deep-content/skill-advisor/actions/workflows/ci.yml) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE) [![Python](https://img.shields.io/badge/python-3.11%2B-blue.svg)](https://www.python.org/downloads/) [![Tests](https://img.shields.io/badge/tests-542%20passing-brightgreen.svg)](#development) [![Hook latency](https://img.shields.io/badge/latency-~0.3s%20warm-success.svg)](#latency-and-the-two-matcher-modes) [![PRs welcome](https://img.shields.io/badge/PRs-welcome-brightgreen.svg)](#contributing)

[**Quickstart**](#quickstart) • [**How it works**](#how-the-advisor-works) • [**Configuration**](#configuration-reference) • [**Architecture**](#architecture-deep-dive) • [**Lifecycle mode**](#lifecycle-mode) • [**Effort signalling**](#effort-signalling) • [**Telemetry**](#telemetry-and-usage-reports)

Overview

`skill-advisor` is a silent `UserPromptSubmit` hook for Claude Code. It inspects every prompt you type, matches it against your installed skill, subagent, and slash-command library, and injects the best picks into the model's context — so the right skill fires automatically, without you lifting a finger.

The advisor is **bring-your-own-library**: it reads skills from `~/.claude/skills/` (plus plugin-marketplace skills and any extra roots you configure) at runtime. This repo ships the runtime only; you supply the skills.

No API keys are ever introduced. All LLM calls route through `claude -p` subprocesses, and embeddings run locally via [`fastembed`](https://github.com/qdrant/fastembed).