CarbeneAI

Specter — Security skill for Claude Code

Security community

Real-time SIEM security dashboard with AI-powered alert analysis.

How to install Specter

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

What Specter does

Real-time SIEM security dashboard with AI-powered alert analysis. Connects to Wazuh & Suricata.

Alternatives in Security

  • Google Workspace Alert Center — Manage security alerts 21.6k ★
  • Security Guardian — Comprehensive security analysis for RTK CLI tool, focusing on command injection, shell escaping, hook security 11.9k ★
  • Deepsec — Deepsec is a security harness for finding vulnerabilities in your codebase powered by coding agents 7.8k ★

README

Specter Banner

Specter

**Real-Time SIEM Dashboard with AI-Powered Alert Analysis**

[![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE) [![Bun](https://img.shields.io/badge/Runtime-Bun-black)](https://bun.sh) [![Vue 3](https://img.shields.io/badge/Frontend-Vue%203-42b883)](https://vuejs.org)

Specter is a real-time security dashboard that connects to your Wazuh SIEM and Suricata IDS. It streams live alerts with severity-based color coding and includes an AI chat panel (Claude) that can autonomously search your historical alerts to provide context-aware threat analysis.

About CarbeneAI

Specter is one of four open-source security tools published by CarbeneAI, alongside [Harbinger](https://github.com/CarbeneAI/Harbinger) (AI threat intel), [Talon](https://github.com/CarbeneAI/Talon) (AI-assisted offensive workflows), and [Forge](https://github.com/CarbeneAI/Forge) (AI operations platform).

CarbeneAI builds open-source tooling for security operations and teaches practitioners to build their own. These repos are the working reference.

Use this repo. Fork it. Improve it. More at [carbene.ai](https://carbene.ai)

Features

  • Deterministic Scorer - Rules-only pre-triage that scores every alert 0-100 and bands it critical/high/medium/low/noise before any LLM is called. No API key, no network call, so it keeps working when the AI chat is unconfigured. Every weight lives in one documented constant.
  • Investigation Ledger - Every AI investigation persisted as a replayable record: system prompt, each tool call and its result, final analysis, token counts, timing. Addressable by permalink. The receipt that makes autonomous response auditable after the fact.
  • Live Alert Streaming - WebSocket-based real-time alert feed from Wazuh Indexer (polls every 30s)
  • Severity Color Coding - CarbeneAI dark theme with Critical/High/Medium/Low visual hierarchy