- 12-stage automated "kill chain" workflow: Reconnaissance, Vulnerability Analysis, Deep & Zero-Day Analysis, API Vulnerability Scan, Database Vulnerability Scan, Registration & Account Provisioning, Authentication & Login, Session Management, Authorization & Access Control, Business Logic, Exploitation Prep, and Post-Exploitation Review
- Flask-based web app with live, real-time output streaming (Server-Sent Events) to a browser frontend
- Port and service scanning via nmap / masscan (full TCP sweep + targeted service/version detection)
- UDP service probing (SNMP, DNS, TFTP, NTP, LDAP, etc.)
- DNS enumeration, zone transfer attempts, and DMARC/wildcard checks
- HTTP/S fingerprinting, WAF/CDN detection, and technology stack identification
- SMB/RPC/NetBIOS enumeration (null sessions, share listing, domain/user enumeration)
- WHOIS, ASN, certificate transparency, and OSINT/subdomain discovery
- CVE mapping via nmap vulners/vulscan and Nuclei template scanning
- ExploitDB lookups via searchsploit
- Web directory/file brute-forcing (gobuster/ffuf) and Nikto web audits
- Deep TLS/SSL analysis: protocol and cipher-suite audit via sslscan/testssl, certificate checks, and Heartbleed detection
- SNMP community string probing
- SMTP/LDAP/FTP security audits
- Banner anomaly detection and version-gap/EOL analysis against a built-in version database
- HTTP behaviour anomaly probes (path traversal, verb tampering, host header injection)
- Timing/side-channel probes for blind SQL injection and authentication oracles
- Authentication and session weakness checks (default credentials, cookie flags)
- XSS, SSTI, XXE, and open-redirect injection surface probing
- CMS and framework detection (WordPress, Drupal, Joomla, Laravel, Django, Rails)
- OWASP API Top-10 style testing: endpoint discovery, BOLA/IDOR, broken authentication, injection, SSRF, mass assignment, and misconfiguration checks
- Database vulnerability scanning across MySQL, PostgreSQL, MongoDB, Redis, MSSQL, Oracle, Cassandra, Elasticsearch, InfluxDB, and Neo4j
- Unauthenticated database access and default credential testing
- Automated SQL injection auditing via sqlmap, plus manual and second-order SQLi probes
- NoSQL injection detection
- Database privilege and configuration auditing
- Exploitation preparation: default credential brute-forcing (hydra), Metasploit safe (non-destructive) checks, sqlmap verification, and exploit/CVE cross-referencing
- Post-exploitation review: network segment discovery, privilege escalation vector lookup, high-value service mapping, credential/secret harvesting, and Active Directory enumeration
- Structured findings output with severity ratings, affected paths, and remediation commands
- PDF report generation with proof-of-concept and code-fix remediation examples
- Command injection safeguards via target input validation and sanitization
- Optional API token authentication (
PWM_API_TOKEN): header/query token for scripts, and a/authpage that stores the token in an HttpOnly cookie so the bundled UI and its EventSource stream work without custom headers
git clone https://github.com/spidermalops/PWM.gitcd PWMLinux / macOS
python3 -m venv venv && source venv/bin/activateWindows
python -m venv venv
source venv\Scripts\activateLinux / macOS
python3 app.pyWindows
python app.pySeven standalone audit suites cover the "no fabricated results" guards:
| Suite | Covers |
|---|---|
test_engine_fixes.py |
port-state parsing, signature false positives, JWT header decode, evidence URLs, stage sequencing |
test_parser_false_positives.py |
negative/positive tool-output fixtures for every substage parser |
test_templates.py |
probe time bounds, template↔parser contracts, timeout-vs-delay handling |
test_session_checks.py |
session-id and cookie/SameSite evidence rules |
test_inprocess_probes.py |
in-process auth/registration probes against a correct vs. flawed app |
test_live_parsers.py |
runs the real API/CMS/DB commands against the fixture and dispatches output to the real parsers |
test_report_status.py |
report metrics and exit-status truth |
# all suites (use the venv interpreter that has requirements installed)
venv/bin/python3 tests/run_all.py
# one suite only
venv/bin/python3 tests/run_all.py templatesEach suite runs in-process and needs no server or external target. Two of them
start the bundled fixture tests/fixtures/vuln_app.py: test_inprocess_probes.py
on 127.0.0.1:8099/8100 and test_live_parsers.py on 127.0.0.1:8110/8111. The
fixture has a secure mode (correct behaviour — a correct tool must stay
silent) and a vuln mode (deliberately flawed — detections must still fire).
It serves the login/registration surface, an API surface (users, admin, config,
search/render, fetch/proxy, swagger, actuator, GraphQL) and CMS fingerprints
(WordPress, Drupal, Joomla), and in vuln mode also starts database-service
stubs on their real ports (Redis 6379, InfluxDB 8086, Elasticsearch 9200,
CouchDB 5984) so the DB probes have a live target. Set FIXTURE_SERVICES=0 to
skip those stubs, FIXTURE_MODE=vuln|secure to choose the mode, and
PWM_FIXTURE_VULN_PORT / PWM_FIXTURE_SEC_PORT to move the web listeners.
test_live_parsers.py therefore needs curl, nmap and redis-cli (and uses
wpscan when present); substages whose tools are missing are reported as
skipped rather than passing. A suite exits non-zero on failure and the runner
exits non-zero if any suite fails.
