How eng-accelerator/outskill-ai-lab works
Outskill AI Lab is a portfolio of five independent, multi-agent products, each with a FastAPI/React web application and an agent workflow. Users submit a task or query through the app (or a project’s CLI), the workflow runs specialized stages, and the result is streamed or returned as a report. The graph preserves the README-documented domain stages and simulator/API capabilities. Source excerpts are incomplete, so web-to-runner wiring and unsampled integrations are shown only where documented, not inferred from directory structure.
Main entry points: aiops_incident_response_agent/main.py Response Runner · cybersecurity_threat_detection_agent/main.py Detection Runner · customer_support_agent/main.py Support Runner · src Incident UI
Components 34
Incident Response
- Incident UI
projects/aiops_incident_response_agent/frontend/src - Incident API
projects/aiops_incident_response_agent/api - Response Runner
projects/aiops_incident_response_agent/main.py - Observability Scenarios
projects/aiops_incident_response_agent/simulators - Triage
projects/aiops_incident_response_agent/agents/triage.py - Analysis and RCA
projects/aiops_incident_response_agent/agents - Remediation and Report
projects/aiops_incident_response_agent/agents
Threat Detection
- Threat UI
projects/cybersecurity_threat_detection_agent/frontend/src - Threat API
projects/cybersecurity_threat_detection_agent/api - Detection Runner
projects/cybersecurity_threat_detection_agent/main.py - Security Scenarios
projects/cybersecurity_threat_detection_agent/simulators - Alert Intake
projects/cybersecurity_threat_detection_agent/agents/alert_intake.py - Threat Analysis and Intel
projects/cybersecurity_threat_detection_agent/agents - Containment and Report
projects/cybersecurity_threat_detection_agent/agents
Customer Support
- Support UI
projects/customer_support_agent/frontend/src - Support API
projects/customer_support_agent/api - Support Runner
projects/customer_support_agent/main.py - Support Scenarios
projects/customer_support_agent/simulators - Intent Routing
projects/customer_support_agent/agents/intake_router.py - Support Specialists
projects/customer_support_agent/agents - Escalation and Resolution
projects/customer_support_agent/agents
Deep Research
- Research UI
projects/deep_research_agent/frontend/src - Research API
projects/deep_research_agent/api - Research Runner
projects/deep_research_agent/main.py - Research Planning
projects/deep_research_agent/agents/research_planner.py - Search and Extraction
projects/deep_research_agent/agents - Synthesis and Report
projects/deep_research_agent/agents/synthesizer.py
Browser Automation
- Browser UI
projects/browser_automation_agent/frontend/src - Automation API
projects/browser_automation_agent/api - Browser Runner
projects/browser_automation_agent/main.py - Task Planning
projects/browser_automation_agent/agents/task_planner.py - Navigation and Interaction
projects/browser_automation_agent/agents - Validation and Report
projects/browser_automation_agent/agents
Other components
- User
Connections 38
- User → Incident UI: submits incident
README.md - Incident UI → Incident API: sends requests · Inferred; not shown in the code that was read.
- Incident API → Response Runner: invokes workflow · Inferred; not shown in the code that was read.
projects/aiops_incident_response_agent/api/routers/incident.py - Response Runner → Observability Scenarios: loads scenarios
projects/aiops_incident_response_agent/main.py - Response Runner → Triage: starts pipeline
README.md - Triage → Analysis and RCA: hands off
README.md - Analysis and RCA → Remediation and Report: hands off
README.md - Remediation and Report → User: returns report
README.md - User → Threat UI: submits alert
README.md - Threat UI → Threat API: sends requests · Inferred; not shown in the code that was read.
- Threat API → Detection Runner: invokes workflow · Inferred; not shown in the code that was read.
- Detection Runner → Security Scenarios: loads scenarios
projects/cybersecurity_threat_detection_agent/main.py - Detection Runner → Alert Intake: starts pipeline
README.md - Alert Intake → Threat Analysis and Intel: hands off
README.md - Threat Analysis and Intel → Containment and Report: hands off
README.md - Containment and Report → User: returns report
README.md - User → Support UI: submits issue
README.md - Support UI → Support API: sends requests · Inferred; not shown in the code that was read.
- Support API → Support Runner: invokes workflow · Inferred; not shown in the code that was read.
- Support Runner → Support Scenarios: loads scenarios
projects/customer_support_agent/main.py - Support Runner → Intent Routing: starts pipeline
README.md - Intent Routing → Support Specialists: routes issue
projects/customer_support_agent/agents/intake_router.py - Support Specialists → Escalation and Resolution: routes or escalates
README.md - Escalation and Resolution → User: returns resolution
README.md - User → Research UI: submits query
README.md - Research UI → Research API: sends requests · Inferred; not shown in the code that was read.
- Research API → Research Runner: invokes workflow · Inferred; not shown in the code that was read.
projects/deep_research_agent/api/routers/research.py - Research Runner → Research Planning: starts pipeline
README.md - Research Planning → Search and Extraction: dispatches searches
README.md - Search and Extraction → Synthesis and Report: provides content
README.md - Synthesis and Report → User: returns cited report
README.md - User → Browser UI: submits task
README.md - Browser UI → Automation API: sends requests · Inferred; not shown in the code that was read.
- Automation API → Browser Runner: invokes workflow · Inferred; not shown in the code that was read.
- Browser Runner → Task Planning: starts pipeline
README.md - Task Planning → Navigation and Interaction: hands off
README.md - Navigation and Interaction → Validation and Report: hands off
README.md - Validation and Report → User: returns result
README.md