Back to skill

Security audit

Financial Investment Banking

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed investment-banking workflow bundle; it handles sensitive financial work but does not show hidden, destructive, or deceptive behavior.

Install only if you are comfortable letting the agent work with confidential deal materials, financial models, templates, and approved market-data sources. Confirm any proprietary-data access, local template inspection, and file creation before use, especially for live M&A matters.

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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This workflow explicitly instructs the agent to run a shell command to enumerate local skills and templates. Even though the stated goal is presentation creation, filesystem inspection expands the skill's authority and can reveal environmental details or be adapted into broader local discovery behavior if the command pattern is reused elsewhere.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The buyer-list description triggers on phrases like "who would buy this," which can occur in ordinary discussion and may invoke the skill outside a clearly bounded M&A context. The file provides positive examples but no exclusion conditions or negative examples to narrow when these triggers should and should not activate.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

Phrases such as "where are we on" and "process update" are broad everyday language and do not clearly limit invocation to live investment-banking deal management. Without stronger context bounds, the skill could match unrelated status-check conversations.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
82% confidence
Finding

The skill description says it creates professional strip profiles, but the workflow mandates pulling data from multiple external systems such as BamSEC, SEC EDGAR, Bloomberg, FactSet, CapIQ, and recent news. That is a materially broader capability—external market intelligence gathering and proprietary data access—than simple slide/profile creation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

Natural-language headings and instructions switch to Chinese in sections such as "来源" and "底层技能" while the rest of the skill is primarily English. This imposes a locale/language assumption without explicit user opt-in or justification for a bilingual or Chinese-specific audience.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.