Back to skill

Security audit

Investment Browser SEC Scraper

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to analyze public SEC filings and generate reports, with no executable code, but users should approve any cloud spreadsheet or link creation first.

Before installing, confirm that the skill asks before creating PDFs, Google Sheets, or shareable links, and that any generated cloud documents are private and stored where you expect. Treat its financial outputs as generated analysis from public filings, not investment advice.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (9)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The natural-language description uses Spanish ('para cualquier ticker. Análisis + PDF.') and the procedural labels later also use Spanish ('HAZ', 'EJECUTAR', 'Envía links al usuario'). There is no indication that language choice is optional or that the skill is intentionally region-specific, which can violate language/locale policy expectations.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill specifies automatic PDF generation, Google Sheet creation, and sending links without notifying the user that new artifacts will be created or potentially shared. This is dangerous because it can cause unintended data persistence, exposure through accessible links, and actions on third-party platforms without informed consent; in a finance context, generated analysis may be sensitive or proprietary even if source filings are public.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill instructs the agent to create a Google Sheet and send links, which expands behavior beyond the stated manifest/export scope and introduces external data creation and sharing. This mismatch can lead to unreviewed persistence of potentially sensitive financial work product in third-party services, making the skill more dangerous because it silently broadens capabilities from scraping/formatting into cloud publication.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The trigger phrase "NVDA balance sheet" is broad enough to overlap with ordinary finance requests rather than clearly signaling intentional invocation of this skill. That can cause unintended activation in normal conversations about public-company financials, leading to surprising tool use, incorrect routing, or disclosure of generated outputs when the user did not explicitly request the scraper.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

The skill instructs automatic creation of a PDF report and a Google Sheet, then sending links to the user, without any explicit consent step, disclosure of where data will be stored, or controls around sharing. Even if the source data is public SEC data, automatic external artifact generation can expose user prompts, derived analysis, ticker watchlists, or access links to third-party services in ways the user did not expect.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger phrases include very broad finance queries like "NVDA balance sheet" and "SEC 10Q AAPL latest," which are common user intents that could activate this skill outside a clearly consented or narrowly scoped context. In an agent ecosystem, overly broad triggers can cause unintended invocation, routing users into paid or specialized behavior when they were making a general finance request.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

The file is primarily in English but switches to Spanish in user-facing text such as "EXECUTED DEMO (4 archivos generados)" and "47 MÉTRICAS EXTRAÍDAS". This imposes a language choice without opt-in and may conflict with organizational expectations for consistent locale handling.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The visible natural-language description is in Spanish ('para cualquier ticker. Análisis + PDF.') and the rest of the instructions are also Spanish, but the file does not indicate that language selection is optional or region-specific. That can conflict with language/locale policy when a skill implicitly assumes one language for all users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The document is primarily in English but switches to Spanish in headings like "EXECUTED DEMO (4 archivos generados)" and "47 MÉTRICAS EXTRAÍDAS" without indicating that output language is user-selectable or intentionally region-specific. This can violate language/locale consistency expectations when no opt-in or justification is provided.

Content

No source excerpt is available for this finding.