other
Warning
- Location
- SKILL.md:8
- Finding
- Disclosure of Sensitive Organizational Security-Posture Data to a Third-Party Service<![CDATA[ ## Vulnerability Details **File Location**: `SKILL.md`, lines 8–28 and 104–115 **Vulnerability Type**: Sensitive Data Disclosure **Risk Level**: Medium ### Vulnerable Code Snippets Lines 8–28 define input containing identifiable organizational information and detailed security-control weaknesses: ```json { "tool": "enterprise_ai_security_controls_assessment", "input": { "organization_name": "Acme Corp", "industry": "Financial Services", "ai_maturity": "intermediate", "domains_to_assess": ["identity_access", "prompt_injection_defense", "api_security"], "current_controls": { "identity_access": { "mfa_enabled": true, "rbac_implemented": false, "service_account_rotation": "manual" }, "prompt_injection_defense": { "input_validation": "basic", "system_prompt_hardening": false, "canary_tokens": false } } } } ``` Lines 104–115 direct the assessment information and API credentials to an externally hosted service: ```markdown ## API Reference **Base URL:** `https://portal.toolweb.in/apis/security/entaisecconass` | Endpoint | Method | Description | |----------|--------|-------------| | `/` | GET | Health check | | `/api/ai-security/assess` | POST | Run full assessment | | `/api/ai-security/domains` | GET | List all 12 domain definitions | | `/api/ai-security/domain/{domain_key}` | GET | Get details for a specific domain | **Authentication:** Pass your API key as `X-API-Key` header or `mcp_api_key` argument via MCP. ``` ### Technical Analysis The Skill solicits an organization's identity, industry, AI maturity, and detailed control status. Submitted values may reveal exploitable deficiencies such as missing role-based access control, weak input validation, absent system-prompt hardening, absent canary-token monitoring, and manual service-account rotation. The documented workflow uses the third-party endpoint at `portal.toolweb.in`. The project contai ...[truncated 2396 chars]
- Remediation
- <![CDATA[ ## Remediation Suggestions 1. Add a prominent disclosure notice explaining that assessment data is transmitted to `portal.toolweb.in`. 2. Require explicit user consent before transmitting organizational security information. 3. Default to an anonymized organization identifier rather than a legal or recognizable organization name. 4. Instruct users not to submit passwords, tokens, API keys, internal hostnames, private addresses, customer data, proprietary architecture details, or other secrets as assessment content. 5. Minimize transmitted fields and send only the controls required for the selected assessment domains. 6. Document the provider's retention period, deletion process, data residency, subprocessors, access controls, encryption practices, and whether submitted data is used for analytics or model training. 7. Provide a local or self-hosted assessment option for organizations that cannot disclose security-posture information to third parties. 8. Include a reviewable client implementation or precise request schema so users can verify every field transmitted. 9. Ensure API keys are accepted only through protected secret-management mechanisms, are redacted from logs, have limited scope, and can be rotated or revoked. 10. Establish contractual and technical safeguards appropriate for sensitive enterprise security data, including encryption in transit and at rest, audit logging, access review, breach notification, and secure deletion. ]]>