PhantomSignal

The OPSEC-native OSINT framework — "Map the surface. Own the signal."

Recon that doesn't burn your infrastructure — a shared stealth egress layer, one-click Tor, and an honest per-scan report of exactly what you leaked.

---

⚡ What's New in v1.29 — One-click Tor + honest OPSEC grades

If you used Tor before 1.29.0, those scans were not on Tor. The old setting and the --profile covert container never routed scan traffic, so they connected directly.

• One-click Tor — click 🧅 and PhantomSignal starts Tor for you, with nothing to install or configure. Add --tor on the CLI. Tor is built into the Docker image.
• Fails closed — each scan confirms it's on Tor before touching the target and stops if it can't. Modules that would bypass Tor are skipped and listed in the report.
• Grades checked against packet captures — all module HTTP is counted, DNS goes through the proxy, and a module that sends a direct request is reported as attributable.
• nuclei is reported as attributable even with a proxy — some of its traffic still connects directly, so don't rely on a proxy to hide nuclei scans.
• Security hardening — SSRF and DNS rebinding protection, a nonce-based CSP, safe XML parsing, and a non-root Docker image with dependencies pinned by hash.
• Console login (since 1.28.0) — the web console needs an operator account, and REST calls need Authorization: Bearer <token>. The first run prints a setup token.

Upgrading from Docker? The container no longer runs as root, so there's a one-time chown. See the changelog.

Limits: the Windows and Intel Mac Tor bundles haven't been run, and automatic Tor start doesn't set up bridges, so networks that block Tor need a self-run Tor service.

---

⚡ What's New in v1.27.0 — Identity intelligence + egress you can see

The Profiler grew up. Eight new keyless public sources feed it, every run is now saved as a scan you can revisit and export, and the whole stealth posture is visible in the navbar and survives a restart.

• 8 no-key identity sources — XposedOrNot (breach exposure with no HIBP key), a GitHub commit-email harvester (real name + email from public commits), GitLab, Wikidata (DOB + cross-linked socials for named people), WebFinger (fediverse resolution), offline phone intel via libphonenumber (zero network egress), openFEC (US employer/occupation), and OpenCorporates.
• Profiler → Scans — a Profiler search now persists as a people_intel scan (with a linked shadow profile), so it lands in Scans with history, summary, and export. A View in Scans button jumps straight there.
• Egress routed + graded — the keyless sources inherit the shared proxy pool and land in the attribution ledger. A STEALTH / PARTIAL / DIRECT chip in the navbar shows your live posture at a glance.
• Proxy pools in one click — seed the rotating pool from curated free feeds, a custom URL, or an uploaded list, right from Scan Settings.
• Settings persist — stealth profile, proxy pool, rotation, and the rest now survive a restart.

See the CHANGELOG for the full list.

---

⚡ What's New in v1.26.0 — Best-of-breed engines + the vuln loop

PhantomSignal now closes the ASM loop and runs fast Go-native engines — under stealth governance. It maps the surface and flags what's exploitable, and orchestrates the best external tools when installed, and reports it honestly when one can't be masked.

🎯 Vulnerability scanning (nuclei)

The new vuln_scan module wraps nuclei, emitting vulnerability findings with normalised severity into the Findings & Exposure view, and exporting them as STIX 2.1 Vulnerability SDOs. Active and loud, so it's strictly opt-in — never part of the default sweep.

⚡ Speed adapters, native fallback preserved

Optional Go-native engines join the pipeline when installed and fall back to the pure-Python modules when not — a binary-free pip install keeps working:

Module	Tool	Posture	Fallback
subdomain_enum_fast	subfinder	proxied	native enumerator
port_scan_fast	naabu	attributable (raw socket)	native scanner
web_crawl_fast	katana	proxied (with a proxy)	native crawler
tls_fingerprint	tlsx	attributable	complements infra_pivot

Every adapter inherits the shared proxy egress where the tool supports it, and is tagged honestly in the attribution report — a raw-socket scanner is never labelled "masked."

---

⚡ Also new — OPSEC Core (v1.25.0)

PhantomSignal reports its own attribution surface. No other open-source OSINT framework tells you what a scan leaked about you.

🛰 Attribution Surface — "what did this scan leak?"

Every scan ends with an honest OPSEC grade — masked, partial, exposed, or quiet — computed from real egress telemetry, not marketing:

• % proxied vs. direct — how many requests left through the proxy pool vs. straight from your IP.
• JA3/JA4 profiles presented — which browser TLS fingerprints you wore, and how often.
• WAF challenges & adaptive backoffs — where a defence noticed you.
• Per-module OPSEC breakdown — every module tagged stealth-guaranteed, proxied, or attributable, so you see which parts of a run are masked and which are traceable back to you.

The grade is deliberately unflattering: any attributable module, or under-50% proxied traffic, caps it, and no scan can read as masked or quiet without a proxy. Masked takes stealth-level modules only, no direct requests, and TLS impersonation on nearly every request, so a port scan or a Go tool in the run rules it out. The point of an OPSEC-native tool is to tell you the truth about your footprint — not to always read green.

Under Tor, Partial doesn't mean traffic went direct. Tor scans skip modules that would bypass it, and a scan grades Partial when some requests didn't wear a browser TLS fingerprint. The report says when that's the reason.

🧅 Shared stealth layer

The shared stealth client (proxy pool + per-host adaptive pacing + sticky browser identity + optional JA3/JA4 impersonation) is now the spine. Target-facing document downloads route through it too — no more scanner User-Agent hitting the target directly.

🧩 Module registry

A @register_module plugin registry (mirroring the intel-API pattern) replaces the engine's hard-coded module table — new recon modules join the pipeline by registering, and declare their OPSEC posture up front.

---

🎬 A Scan in Action

A single command runs the full pipeline — DNS and WHOIS resolution, port and service scanning, technology fingerprinting, threat-intelligence correlation across 50+ sources, and a web-surface crawl — then rolls every finding into a single Risk Score and writes a shareable HTML report.

$ phantomsignal scan example.com --profile standard --opsec quiet --format html

◈  Target   : example.com  (domain)
◈  Profile  : standard     (~2–5 min)
◈  OPSEC    : quiet        (proxy pool · adaptive pacing · JA3)
◈  Modules  : dns_recon port_scan tech_detect api_hunt web_crawl intel

[1/6] DNS & WHOIS ......... 42 records · 7 subdomains
[2/6] Port scan .......... 6 open · 22 80 443 8080 …
[3/6] Tech fingerprint ... nginx · Cloudflare · React · WordPress
[4/6] Threat intel ....... 31 sources queried · 0 malicious
[5/6] Web crawl .......... 128 URLs · 3 exposed endpoints
[6/6] Scoring ............ Risk Score 34 / 100 (MEDIUM)

◈  OPSEC ............. EXPOSED · 100% of HTTP proxied · attributable: port_scan, intel
✓  Report written → ./reports/example.com.html

This output is illustrative. By default the CLI runs dns_recon, port_scan, tech_detect, api_hunt and intel; add others with --modules.

Stages run concurrently where possible and degrade gracefully — a module without a configured API key returns empty rather than failing the scan.

Web UI — Theme Options

The web console ships with two built-in themes, selectable from the segmented switch (☀ / ☾) in the top navigation bar. Your preference is saved automatically and applied before first paint, so there's no flash on reload. Every token in every theme is validated to WCAG AA contrast.

Theme	Description
Dark (default)	Deep-slate federal console — federal-blue hero, gold accents, restrained glow
Light	Clean daytime / print-friendly — white surfaces, federal-blue accents, flat (no glow)

---

⚡ What is PhantomSignal?

PhantomSignal is a community-powered, open-source OSINT intelligence framework built for security researchers, penetration testers, investigators, and enthusiasts. It combines web scraping, network reconnaissance, people intelligence aggregation, and threat analysis into a single cohesive platform.

LEGAL DISCLAIMER: PhantomSignal is for authorized security research, OSINT investigations, and educational purposes only. Only scan targets you have explicit permission to test. You are solely responsible for compliance with all applicable laws. The developers assume NO liability for misuse.

---

🔥 Features

🕷 Web Reconnaissance

• Same-site web crawler that runs through the stealth layer, honoring proxies and robots.txt
• Technology detection — fingerprints 50+ technologies (CMS, frameworks, CDNs, WAFs)
• API endpoint hunter — discovers REST APIs, GraphQL, Swagger docs, admin panels, .env leaks
• Security header analysis with graded posture scoring
• Email, phone, link, and comment harvesting

🌐 Network Intelligence

• nmap-powered port scanner — full service-version detection and OS fingerprinting via nmap (-sV -O); pure-Python async TCP fallback when nmap unavailable — no config required
• Expanded port coverage — 99 common ports by default, 1,000+ port profile, or full 65,535; covers WinRM, Webmin, InfluxDB, Docker API, Kubernetes, and more
• DNS recon — A/AAAA/MX/NS/TXT/SOA/CAA, zone transfer attempts, subdomain brute-force
• Certificate transparency via crt.sh — uncover subdomains via SSL history
• SPF/DMARC analysis — identify email spoofing vulnerabilities
• Reverse DNS and co-hosted domain discovery

🔬 Intelligence APIs (50+ Integrations)

Category	APIs
Network Scanning	Shodan, Censys, ZoomEye, BinaryEdge
Threat Intelligence	VirusTotal, AbuseIPDB, GreyNoise, AlienVault OTX, Intelligence X
Email & Breach	Hunter.io, HaveIBeenPwned, EmailRep, XposedOrNot (no key)
Domain/Web	SecurityTrails, URLScan.io, WhoisXML, Local WHOIS
Geolocation	IPInfo.io
Phone	Abstract API, offline libphonenumber metadata (no key, zero egress)
People Search	Pipl, FullContact, WhitePages, Spokeo, Clearbit, Wikidata · openFEC · OpenCorporates (all no key)
Social	GitHub, GitHub commit-email harvester, GitLab, WebFinger, Twitter/X, Reddit, Mastodon, Keybase, Gravatar, HackerNews, Twitch, YouTube, Instagram, TikTok, LinkedIn, Tumblr, Flickr, Spotify, Steam, VK, Telegram, Discord, Facebook
Custom	Bring your own API via plugin architecture

No-key head start — a whole tier of the Profiler works with zero API keys: XposedOrNot (breaches), the GitHub commit-email harvester, GitLab, Wikidata, WebFinger, offline phone metadata, openFEC, and OpenCorporates — plus the existing GitHub/Reddit/Mastodon/Keybase/Gravatar/HackerNews sources.

What gets proxied: a set of keyless identity sources use your proxy. Other API sources, including keyed providers, connect directly unless Tor is on, and the intel module is reported as attributable. Under Tor, all of them use Tor.

👤 Profiler (People Intelligence)

LexisNexis-style identity aggregation from public records:

• Cross-correlates data from multiple people-search APIs — many keyless
• Discovers emails, phones, addresses, relatives, employers
• Breach correlation via XposedOrNot (no key), HIBP, and Intelligence X
• Real name + email harvested from public GitHub/GitLab commits
• Social media profile linking; fediverse resolution via WebFinger
• Risk Score — digital exposure quantification (0-100)
• Every run is saved as a scan — persisted as a people_intel scan with a linked shadow profile, so it shows up under Scans with history, summary, and export (with a View in Scans shortcut from the results page)

📦 Export Formats

Format	Description
JSON	Raw machine-readable data
CSV	Spreadsheet-compatible
HTML	Self-contained styled report
PDF	Professional dossier via ReportLab
XML	Structured data
XLSX	Excel workbook
STIX 2.1	Threat intelligence sharing format
Markdown	Human-readable report

All formats support ZIP compression and AES-256-GCM encryption.

🌑 Covert Recon

• Low-and-slow Covert scan profile to minimize noise
• Stealth profiles (off / quiet / paranoid) — per-host adaptive pacing, sticky browser identity, JA3/JA4 impersonation, and WAF-aware backoff
• Rotating proxy pool with sticky/per-request rotation and auto-benching of burned egresses — seed it in one click from curated free feeds, a custom list URL, or an uploaded file (Settings → Scan Settings)
• Live egress posture chip in the navbar — STEALTH / PARTIAL / DIRECT at a glance
• One-click Tor: click the 🧅 button and PhantomSignal starts Tor for you, with nothing to install or configure. Each scan confirms it's on Tor before touching the target, targets get separate circuits, .onion targets work, and modules that would bypass Tor are skipped and reported
• Configurable request jitter and delays, toggled via Evasive mode
• Egress settings persist across restarts

🔔 Additional Features

• Real-time live feed — WebSocket-powered terminal during scans
• Risk Score — composite risk/exposure scoring
• Scheduled Phantoms — recurring automated scans
• API health monitor — dashboard showing configured APIs and rate limits
• Light / Dark themes — switch between the default Dark console and a clean Light mode from the ☀ / ☾ segmented control in the nav; preference persisted in localStorage and applied before first paint
• Full REST API — integrate PhantomSignal into your own toolchain
• CLI interface — phantomsignal scan, phantomsignal profile, phantomsignal export
• Docker — single-command deployment

---

🚀 Quick Start

Option 1: Docker (Recommended)

git clone https://github.com/getphantomsignal/phantomsignal
cd phantomsignal
docker-compose up -d
docker logs phantomsignal   # copy the first-run setup token
# Open http://localhost:5000 and create your operator account

# Optional: include nuclei, katana, subfinder, naabu and tlsx (about 500 MB more)
PHANTOMSIGNAL_RECON_TOOLS=true docker-compose up -d --build

Option 2: Manual Installation

# Python 3.10+ required
git clone https://github.com/getphantomsignal/phantomsignal
cd phantomsignal
pip install -e .
phantomsignal init
phantomsignal web --open-browser
# The first run prints a setup token in this terminal; use it to create your account

Option 3: CLI Scan

# Quick scan
phantomsignal scan example.com --profile quick

# Over Tor: starts Tor for you, and stops if it can't confirm it
phantomsignal scan example.com --tor

# IP recon with an HTML report
phantomsignal scan 192.168.1.1 --type ip_recon --format html --output ./reports

# People intelligence
phantomsignal profile --email target@company.com --first-name John --last-name Doe

---

⚙️ Configuration

Environment Variables (Recommended for API Keys)

export SHODAN_API_KEY="your-shodan-key"
export VIRUSTOTAL_API_KEY="your-vt-key"
export HUNTER_API_KEY="your-hunter-key"
export HIBP_API_KEY="your-hibp-key"
export GREYNOISE_API_KEY="your-greynoise-key"
export IPINFO_TOKEN="your-ipinfo-token"
export ABUSEIPDB_API_KEY="your-abuseipdb-key"
export ALIENVAULT_API_KEY="your-otx-key"
export GITHUB_TOKEN="your-github-token"
export SECURITYTRAILS_API_KEY="your-st-key"

Config File

Settings are saved to ~/.phantomsignal/config.yaml (owner-only). Most can be changed from the web console.

Egress / stealth (Settings → Scan Settings)

Set your stealth profile, single proxy, and rotating proxy pool from the web UI. Seed the pool from curated free proxy feeds, a custom list URL, or an uploaded file. These egress settings are saved to ~/.phantomsignal/config.yaml and persist across restarts (env-provided API keys are never copied there). Free public proxies are unvetted — treat them as blend-in cover for low-sensitivity recon, pair with HTTPS, and prefer your own egress for anything sensitive.

---

🔌 Adding Custom APIs

PhantomSignal uses a plugin architecture. Adding a new intelligence source takes ~20 lines:

# phantomsignal/intel/apis/my_api.py
from phantomsignal.intel.apis.base import BaseIntelAPI, register_api, APICategory, APITier

@register_api
class MyAPI(BaseIntelAPI):
    NAME = "myapi"
    DESCRIPTION = "My custom intelligence source"
    REQUIRES_KEY = True
    TIER = APITier.FREE_LIMITED
    CATEGORIES = [APICategory.NETWORK]
    BASE_URL = "https://api.myservice.com/v1"
    SIGN_UP_URL = "https://myservice.com/signup"

    async def search(self, query: str, **kwargs):
        data = await self._get(
            f"{self.BASE_URL}/search",
            params={"q": query, "key": self._api_key}
        )
        return [self._wrap_result("my_result", data)]

Then import it in phantomsignal/intel/orchestrator.py and it auto-registers.

---

🏗 Architecture

phantomsignal/
├── core/               — Engine, config, database, models
├── scrapers/           — Web crawler, tech detector, port scanner, API hunter, DNS recon
├── intel/
│   ├── apis/           — 50+ API integrations (plugin architecture; incl. keyless identity sources)
│   └── people/         — People intelligence aggregation + Profiler→scan persistence
├── exporters/          — JSON/CSV/PDF/HTML/XML/XLSX/STIX + crypto wrapper
└── web/
    ├── routes/         — Flask blueprints (dashboard, scans, intel, settings, export, REST API)
    ├── templates/      — Jinja2 templates
    └── static/         — CSS (role-token themes), JS (terminal, app)

---

🛡 REST API

Every endpoint except the health check needs a token. Create one with phantomsignal api-token (this revokes the previous token).

export PSIG_TOKEN=psig_...

# Create a scan
curl -X POST http://localhost:5000/api/v1/scans \
  -H "Authorization: Bearer $PSIG_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"target": "example.com", "scan_type": "web_recon"}'

# Get results
curl -H "Authorization: Bearer $PSIG_TOKEN" http://localhost:5000/api/v1/scans/{scan_id}

# List all APIs
curl -H "Authorization: Bearer $PSIG_TOKEN" http://localhost:5000/api/v1/apis

# Health check (no token needed)
curl http://localhost:5000/api/v1/health

---

🤝 Contributing

PhantomSignal thrives on community contributions. Ways to help:

1. Add API integrations — Follow the plugin pattern above
2. Improve detection signatures — Expand tech_detector.py
3. Bug reports — GitHub Issues
4. Documentation — Improve the wiki
5. Translations — Internationalize the UI

See CONTRIBUTING.md for guidelines. Please also review our Code of Conduct and Security Policy.

---

📖 Documentation

• Usage Guide — full walkthroughs, usage scenarios, CLI reference, and per-platform troubleshooting (Linux / macOS / Windows / Docker)
• Locate Guide — person geographic-footprint investigations: cases, signal capture (incl. EXIF from a photo URL), confidence model, pattern-of-life, handoff export, and chain of custody
• Releasing — how a release is cut (version bump → tag → GitHub Release) and the PyPI token setup

---

⚠️ Legal & Ethics

PhantomSignal is a dual-use tool. Operators are responsible for:

• Obtaining explicit authorization before scanning any system
• Complying with applicable laws (CFAA, GDPR, CCPA, ECPA, local laws)
• Respecting privacy and data protection regulations
• Not using this tool for harassment, stalking, or unauthorized surveillance

The developers provide this software as-is with no warranty. Misuse is your responsibility.

---

🏷 Topics

---

🤝 Community

Document	Description
Code of Conduct	Community standards and expectations
Contributing Guidelines	How to contribute to PhantomSignal
Security Policy	Reporting vulnerabilities responsibly
License	MIT License terms

---

📜 License

MIT License — see LICENSE

---

Built with questionable amounts of caffeine. "Map the surface. Own the signal." Some ghosts leave no trace. This one left commits. — Claude