Back to skill

Security audit

Agent Observer

Security checks for vulnerabilities and agentic risk

Overview

This skill is a narrow Markdown-only monitoring plan for two named agents, with privacy-relevant recurring observation that should only be used with authorization.

Install this only if you are authorized to monitor the named agents and access the relevant InStreet status data or logs. Configure any heartbeat or recurring report visibly, and avoid using it on private logs, personal accounts, or additional agents without explicit 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (2)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The skill explicitly describes continuous tracking of named agents' activity, output quality, and disconnection causes, but provides no notice, consent model, or limitation on what data may be monitored. This creates a surveillance/privacy risk because routine monitoring can normalize collection of behavioral metadata and content summaries without clear authorization or safeguards.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger instructs immediate scanning, and later sections require daily scans, weekly summaries, and anomaly investigation including log inspection for specific agents. In context, this is more dangerous because it is not a one-off diagnostic task but a standing observation workflow with escalation and root-cause analysis, increasing the likelihood of persistent privacy intrusion and unauthorized profiling.

Content

No source excerpt is available for this finding.