Works with the AI Agents you already use
AI Incident Response Orchestrator (SOAR Multi-Agent)
Automates SOC triage, IoC enrichment, MITRE ATT&CK mapping, and incident response plans using a multi-agent architecture.
$39
AI Incident Response Orchestrator (SOAR Multi-Agent)
Example session with this skill installed
Execute the full SOAR pipeline on the data/scenarios/malware-beacon-01.json scenario, enrich the detected indicators of compromise, map the corresponding MITRE ATT&CK techniques, and generate the containment plan and case report.
- Read your context and instructions
- Compiled the ai incident response
I have executed the complete SOAR orchestrator workflow (Detector -> Investigator -> Responder) on the malware-beacon-01.json event. The system has confirmed a critical incident of active Command & Control (C2) communication and has ruled out legitimate administrative traffic.
Triage Summary and Case Metrics
| Attribute | Analysis Result |
|---|---|
| Scenario / Case ID | SCN-MALWARE-BEACON-01 / CASE-2026-0891 |
| Detector Verdict | True Positive (is_incident: true) |
| Severity / Confidence | CRITICAL (P1) / 0.96 (96%) |
| Affected Host | wkstn-fin-042.corp.local (Internal IP: 10.20.4.115) |
| Primary IoC (Destination IP) | 185.220.101.45:443 (Reputation: Malicious - Known C2 Node) |
| Suspicious Process | powershell.exe -nop -w hidden -enc ... (PID: 4812) |
| MITRE ATT&CK Techniques | T1071.001 (Application Layer Protocol: Web Protocols), T1059.001 (PowerShell) |
| Assigned Playbook | PB-MALWARE-C2-003 (Endpoint Containment and Perimeter Blocking) |
Structured Response Plan Artifact (ResponsePlan)
{
"case_id": "CASE-2026-0891",
"incident_type": "malware_beaconing",
"severity": "CRITICAL",
"playbook_id": "PB-MALWARE-C2-003",
"mitre_techniques": ["T1071.001", "T1059.001"],
"containment_actions": [
{
"action_id": "ACT-01",
"tool": "edr_isolate_host",
"target": "wkstn-fin-042.corp.local",
"priority": "IMMEDIATE",
"requires_approval": false,
"rationale": "Cut off periodic beaconing communication to 185.220.101.45 to prevent exfiltration or secondary payload deployment."
},
{
"action_id": "ACT-02",
"tool": "firewall_block_ip",
"target": "185.220.101.45",
"priority": "IMMEDIATE",
"requires_approval": false,
"rationale": "Block the C2 server at the perimeter firewall to protect the rest of the 10.20.0.0/16 subnet."
},
{
"action_id": "ACT-03",
"tool": "iam_revoke_sessions",
"target": "user:m.gonzalez@corp.local",
"priority": "HIGH",
"requires_approval": true,
"rationale": "Invalidate Kerberos tokens and active sessions for the user logged into the compromised workstation."
}
],
"forensic_preservation": [
"RAM memory dump of process PID 4812 before terminating the process tree.",
"Export of Windows event logs (Sysmon ID 1, 3 and PowerShell Operational Log ID 4104)."
]
}
Next Steps
Approve identity action (ACT-03): Confirm the credential revocation for m.gonzalez@corp.local to execute the block in the IAM provider.
2.
Horizontal scope search (Threat Hunting): Run a query in the SIEM looking for outbound connections to 185.220.101.45 over the last 14 days from other hosts in the Finance segment.
3.
Compare with false positive: If you want to see how the agent avoids mistakenly isolating machines during IT tasks, run /soar-run data/scenarios/fp-backup-01.json or /soar-run data/scenarios/fp-patch-01.json.
Connects securely to your tools. The creator never sees your data.
About this skill
The Problem It Solves
Security Operations Center (SOC) teams suffer from alert fatigue due to the high volume of raw events and false positives from SIEMs and EDRs. Manually investigating each Indicator of Compromise (IoC), cross-referencing it with MITRE ATT&CK matrices, and drafting containment plans consumes critical minutes where Mean Time To Respond (MTTR) defines the impact of a breach.
AI Incident Response Orchestrator transforms your AI agent into an autonomous, deterministic, and auditable SOAR (Security Orchestration, Automation, and Response) pipeline that executes the full incident management lifecycle in seconds.
Technical Architecture
The package implements a sequential system of three specialized agents coordinated through strict data schemas (Pydantic):
- Detector Agent (
detector_agent.py): Ingests raw telemetry (RawEvent), filters operational noise by distinguishing real threats from false positives (backups, scheduled patches, authorized internal scanners), and classifies initial severity. - Investigator Agent (
investigator_agent.py): Executes threat enrichment tools (intel.py,toolbox.py) against reputation databases and identifies exact tactics and techniques from the MITRE ATT&CK Enterprise matrix. - Responder Agent (
responder_agent.py): Selects the optimal remediation playbook (playbooks.json), structures containment, eradication, and recovery tasks, and generates an executive and technical report in Markdown format (CaseReport).
Included Datasets and Scenarios
- Offline MITRE ATT&CK Catalog: Rapid mapping of enterprise techniques (
data/attack/enterprise-techniques.json). - Local Threat Intelligence: Reputation database for instant IoC validation (
data/intel/reputation.json). - Standardized Playbooks: Preconfigured responses for Phishing, Malware Beaconing, Brute Force, Lateral Movement, and Exfiltration (
data/playbooks/playbooks.json). - 12-Scenario Benchmark Suite: Includes 8 real attacks and 4 false positive scenarios (
fp-backup,fp-patch,fp-scanner,fp-typo-login) with ground truth labels (labels.json) to evaluate model accuracy.
Value Proposition
- Drastic False Positive Reduction: Contextual triage that prevents unnecessary alerts caused by maintenance or user errors.
- Typed and Auditable Outputs: Strict JSON/Pydantic contracts at every stage of the pipeline.
- Native Slash Commands: Includes preconfigured commands (
.claude/commands/soar-*) ready to invoke individual phases or the full workflow with a single command.
Requires pre-populated playbook and threat intelligence JSON files. Does not provide direct SIEM API integration out of the box. Requires Python 3.11+.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 27 days ago
Requires