Back to skill

Security audit

GovGazette Exclusion Review

Security checks for vulnerabilities and agentic risk

Overview

The skill performs clearly scoped exclusion-record lookups through GovGazette, with appropriate limits on sensitive inputs and uncertain matches.

Lookups send supplied names and identifiers to GovGazette. Submit only information you are authorized to share, keep confidential material out, and verify official records before relying on a possible match.

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
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (1)

The workflow sends user-supplied names and identifiers to GovGazette without a privacy disclosure.

Medium
Category
Output Handling
Confidence
90% confidence
Finding

The skill instructs the agent to collect names and identifiers and submit them to GovGazette through check_exclusions, but does not warn that this information is transmitted to a third-party service. The instruction to exclude credentials and confidential bid material does not disclose the privacy implications of sending identity information.

Content

No source excerpt is available for this finding.