Back to skill

Security audit

Radar CNPJ

Security checks for vulnerabilities and agentic risk

Overview

The skill’s company checks, paid reports, and optional monitoring fit its stated purpose and include spending controls.

Share supplier identifiers only when your organization permits disclosure to Radar CNPJ, and exclude internal notes. Set an authorized spending limit and verify monitoring duration and cancellation options. External pricing and billing pages were inaccessible during review, and the remote service’s implementation and data retention were not verified.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

The supplier-batch workflow sends the user's supplier list to a third-party service without a privacy warning.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The workflow uploads the user's supplier list to Radar CNPJ without requiring notice or checking whether external disclosure is authorized. Although individual CNPJs are public identifiers, their aggregation as a user's supplier portfolio can reveal confidential business relationships; the instruction to use this skill even when the user does not name Radar increases this risk. Existing spending authority does not necessarily authorize disclosure of that portfolio.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 71)May include surrounding context.

After reading the quote and confirming existing spending authority:

sh
curl --fail-with-body --silent --show-error \
  'https://radar-cnpj.com/api/consulta/lote?formato=csv&lingua=pt' \
  -H "X-Credito: $CREDITO" -H 'Content-Type: application/json' \
  -H "Idempotency-Key: $IDEMPOTENCY_KEY" \