explaingit

parsasmahmoudi/dayzero

1PythonAudience · developerComplexity · 4/5ActiveSetup · moderate

TLDR

Agentic bug-bounty command-line tool that chains reconnaissance and scanning steps, stores findings in SQLite, and streams events to a real-time dashboard.

Mindmap

mindmap
  root((dayzero))
    Inputs
      Target domain
      LLM API keys
    Outputs
      SQLite findings
      Real-time dashboard
      Vulnerability reports
    Use Cases
      Authorized recon
      Subdomain enumeration
      Secret scanning
      JS analysis
    Tech Stack
      Python
      SQLite
      Server Sent Events
      ZeroClaw

Things people build with this

USE CASE 1

Run an automated agentic recon and scan on an authorized bounty target

USE CASE 2

Enumerate subdomains via certificate transparency and probe for CORS issues

USE CASE 3

Scan JavaScript files for leaked secrets during a hunt

USE CASE 4

Stream live findings to a dashboard at port 8765 for review

Tech stack

PythonSQLiteSSE

Getting it running

Difficulty · moderate Time to first run · 30min

Ships in Full YOLO mode with safety gates off; only run against authorized targets.

In plain English

DayZero is a command-line and web-dashboard tool for automated bug bounty work, meaning the practice of probing websites for security flaws under a program where the owner has agreed to be tested. The README presents it as an open-source agentic framework built on top of another project called ZeroClaw. The author positions it as a way for a single researcher to run a long chain of reconnaissance and scanning steps without doing each one by hand. The central piece is what the author calls the Hunt Brain. It looks at what early tools have found, such as discovered subdomains, detected technologies, leaked secrets, or vulnerability hits, and then queues follow-up tools accordingly. The README lists twelve integrated tools that the brain can pull from, including certificate transparency lookups for subdomains, JavaScript file analysis, a CORS probe, a parameter fuzzer, SSL and port checks, secret scanning, and a set of heuristic vulnerability checks. Findings are stored in a local SQLite database so they persist across sessions, and a real-time dashboard at port 8765 streams events using server sent events. It can also call out to a language model for reasoning during a mission. The bring your own key setup supports OpenRouter, OpenAI, Anthropic, Google AI, Groq, Ollama for local models, and any OpenAI compatible custom endpoint. The README says keys live only on the user's machine and never in the repository, and dayzero reset-api can wipe stored keys. Installation is a one-line script for Windows PowerShell or Linux and macOS shell, or a manual git clone plus an install command. After install, dayzero hunt --target example.com runs the default agentic hunt, with --quick, --deep, and --full variants for shallower or fuller scans. The README is upfront that it ships in a Full YOLO mode with safety gates off and frames it for authorized lab use. The repository has 1 star and version 3.1.0.

Copy-paste prompts

Prompt 1
Install dayzero with the one-line script and run a quick hunt against my authorized lab target
Prompt 2
Configure OpenRouter as the reasoning provider and confirm the key stays local
Prompt 3
Walk me through what the Hunt Brain queues next after a subdomain discovery
Prompt 4
Help me extend dayzero with a custom tool plugged into the agentic loop
Open on GitHub → Explain another repo

Generated 2026-05-22 · Model: sonnet-4-6 · Verify against the repo before relying on details.