MITRE ATT&CK Technique Mapper

• Outcome: Review
• Latest audit: Sep 12, 2026 · 7:22 AM UTC
• Version: 1.0.0

Overview

The skill does what it says, but it can send sensitive security reports and incident details to an external ToolWeb API without enough consent, redaction, or data-handling guidance.

Before installing, treat this as an external data-processing integration. Only submit reports or logs you are authorized to share with ToolWeb, redact secrets and unnecessary internal details, use a revocable scoped API key, and check the provider's privacy, retention, and deletion terms.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

Findings (1)

other

Warning

• Location: SKILL.md:27
• Finding: Potential Disclosure of Sensitive Security Reports to an External Service<![CDATA[ ## Vulnerability Details **File Location**: `SKILL.md`, lines 27–39 **Vulnerability Type**: External transmission of potentially sensitive files without documented safeguards **Risk Level**: Medium ### Complete Code Snippet ```bash curl -X POST https://portal.toolweb.in/apis/security/mitre-attack-mapper/map-technique \ -H "X-API-Key: YOUR_API_KEY" \ -F "file=@incident_report.pdf" ``` ### Option 3 — Text + File Combined ```bash curl -X POST https://portal.toolweb.in/apis/security/mitre-attack-mapper/map-technique \ -H "X-API-Key: YOUR_API_KEY" \ -F "input=Focus on lateral movement and credential access techniques" \ -F "file=@threat_intel_report.txt" ``` ### Technical Analysis The documented workflow directs users or an AI agent to upload local incident reports and threat-intelligence documents to `portal.toolweb.in`, an external service. Elsewhere in the document, recommended upload types include incident-response reports, SIEM/EDR exports, penetration-test findings, and malware-analysis reports. These files may contain confidential hostnames, internal IP addresses, usernames, vulnerabilities, indicators of compromise, log records, or other incident evidence. The Skill does not require explicit approval immediately before transmission, recommend redaction, restrict eligible local paths, or document the external service's retention and privacy controls. Consequently, an agent following the instructions could transmit sensitive organizational data without the user fully understanding that the complete file leaves the local environment. The request also transmits an API key to the external endpoint as part of normal authentication. This is operationally necessary for the documented API, but it makes the trustworthiness and transport security of the external service relevant to both the uploaded data and credential. ### Attack Path 1. A user asks the agent to map a local incident or threat-intelligence report to MITRE ATT&amp;CK techniq ...[truncated 1418 chars]
• Remediation: <![CDATA[ ## Remediation Suggestions 1. Require explicit, informed user confirmation immediately before transmitting any local file, clearly naming the destination domain and selected file. 2. Add a prominent warning that uploaded data leaves the local environment and may be processed or retained by a third party. 3. Classify and redact reports before upload, removing credentials, access tokens, personal data, internal hostnames, unnecessary IP addresses, and unrelated log content. 4. Restrict uploads to user-approved paths and expected file types; reject implicit selection of arbitrary local files. 5. Apply file-size limits and validate file content rather than relying only on filename extensions. 6. Document the service's privacy policy, retention period, deletion process, subprocessors, geographic processing locations, and access controls. 7. Use narrowly scoped, revocable API keys stored in a secret manager. Never embed real keys in Skill content, command history, logs, or generated output. 8. Provide a local or self-hosted processing option for confidential incident data. 9. Minimize transmitted information by preferring a sanitized text extract over the complete source document whenever possible. 10. Record an auditable consent event containing the destination and file identity, but never the report contents or API key. ]]>

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• YARA Signatures: Malware Match, Webshell Match, Cryptominer Match
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

YARA rule 'info_stealer': Information stealer patterns (credential harvesting, browser data theft) [malware]

High

• Category: YARA Match

Content

": "Monitor for net user, whoami /all, Get-LocalUser execution",
      "mitigations": ["M1028 - Operating System Configuration"]
    },
    {
      "technique_id": "T1003.001",
      "technique_name": "OS Credential Dumping: LSASS Memory",
      "tactic": "Credential Access",
      "confidence": 0.97,
      "detection": "Monitor for lsass.exe memory access (Sysmon Event ID 10); alert on procdump, mimikatz, Task Manager targeting lsass",
      "mitigations": ["M1043 - Credential Access Protection", "M1028 - Operating System Configuration", "M1026 - Privileged Account Management"]
    }
  ],
  "threat_actor_associations": ["APT29", "FIN7", "Lazarus Group"],
  "kill_chain_coverage": ["Initial Access", "Execution", "Discovery", "Credential Access"]
}
```

---

## API Reference

**Base URL:** `https://portal.toolweb.in/apis/security/mitre-attack-mapper`

| Endpoint | Method | Content-Type | Description |
|----------|--------|--------------|-------------|
| `/map-technique` | POST | `multipa

• Confidence: 75% confidence
• Finding: YARA rule matched a known malware signature (reverse shell, backdoor, ransomware, C2 framework, or info stealer).

Missing User Warnings

Medium

• Confidence: 96% confidence
• Finding: The skill explicitly instructs users to submit free-text incident details and upload reports to a third-party API endpoint, but it does not warn that potentially sensitive security data will leave the local environment. This creates a real data exposure and consent risk because users may unknowingly transmit incident reports, logs, or threat intelligence containing confidential internal information to an external service.

Static analysis

No suspicious patterns detected.