Back to skill

Security audit

Crm Agents

Security checks for vulnerabilities and agentic risk

Overview

This skill is a simple, disclosed connector to a third-party CRM agent API, but users should avoid sending confidential business data unless they trust SputnikX.

Install only if you intend to use SputnikX's remote CRM agents. Treat anything placed in the task field as data sent to a third party, and do not include secrets, customer PII, regulated records, or confidential business data without approval.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill description is extremely broad ('Execute 13 CRM AI agents on demand' and 'Business intelligence as a service'), which creates weak activation boundaries and raises the chance the skill is invoked for generic business-analysis prompts. In a user-invocable skill that sends requests to a third-party API, overbroad matching can cause unintended remote execution and data disclosure without the user clearly intending to use this external service.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
84% confidence
Finding

The skill makes an external network call to a non-local third-party domain, which is a real data egress capability even if the listed endpoint only enumerates available skills. In context, this is less severe than task execution, but it still establishes outbound communication to an external operator and should be transparently disclosed and constrained.

Content

Scanner excerpt · SKILL.md (reported line 18)May include surrounding context.

List Available Skills (free)

bash
curl https://sputnikx.xyz/api/v1/agent/skills

Returns: All available agents with descriptions and capabilities.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill instructs users to submit arbitrary 'task' content to a remote third-party API but does not warn that those contents leave the local environment. Because the tasks are framed as CRM, revenue, supply chain, deals, risk, and compliance analysis, users may include highly sensitive commercial or regulated data without informed consent.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The auto-route feature is activated by a vague natural-language task description ('describe what you need'), which delegates service selection to the remote system without clear user consent boundaries. This increases the risk of accidental invocation, unintended transmission of sensitive business context, and opaque execution against a paid external service.

Content

No source excerpt is available for this finding.