How wpsec/artex-pro works
ARTEX-Pro is an AI-assisted autonomous penetration-testing system. An operator configures a task and authorized scope, then the backend’s event-driven engine coordinates planning and worker agents that use tools to explore targets and record assets, evidence, and findings. Users can inspect task activity, traffic, reports, and findings, and initiate vulnerability retests. The README documents these product capabilities; sampled source confirms key backend dependencies, but many route-level connections and unsampled subsystems cannot be verified here, so those links are omitted or marked unverified.
Main entry points: server.go HTTP API · worker.go Worker agents · engine.go Exploration engine · manager.go Task manager
Components 20
User experience
Task execution
- Task manager
server/manager.go - Exploration engine
server/engine.go - Planner
agent/planner.go - Worker agents
agent/worker.go - Agent tools
agent/tools.go - Action guard
guard/guard.go
Security workflow
- Exploration state
db/exploration.go - Asset scope
db/assets.go - Traffic capture
traffic/traffic.go - Finding evidence
evidence/store.go - Findings
db/findings.go - Vulnerability retests
agent/retester.go
Platform services
- MCP and ZAP
cmd/artex/zap.go - LLM services
- PostgreSQL
db/db.go - Notifications
notify/notify.go
Other components
- Operator
Connections 23
- Operator → Web UI: uses
- Web UI → HTTP API: requests API
- HTTP API → Task manager: handles tasks
server/server.go - Task manager → Exploration engine: coordinates tasks
server/manager.go - Exploration engine → Planner: plans work
server/engine.go - Exploration engine → Worker agents: runs workers
server/engine.go - Worker agents → Agent tools: invokes tools
agent/worker.go - Worker agents → Action guard: gates actions
agent/worker.go - Worker agents → Exploration state: records exploration
agent/worker.go - Worker agents → Asset scope: updates assets
agent/worker.go - Agent tools → MCP and ZAP: uses MCP tools
- Planner → LLM services: requests plans
- Worker agents → LLM services: requests completion
- Exploration engine → Exploration state: checks active intent
server/engine.go - Task manager → PostgreSQL: connects PostgreSQL
server/manager.go - HTTP API → Findings: serves findings
server/server.go - HTTP API → Exploration state: reads graph
server/server.go - Traffic capture → PostgreSQL: indexes exchanges
traffic/traffic.go - Finding evidence → Traffic capture: snapshots traffic
evidence/store.go - Finding evidence → Findings: binds evidence
evidence/store.go - HTTP API → Vulnerability retests: starts retest
- HTTP API → Reports: provides reports
server/server.go - HTTP API → Notifications: configures delivery