Market Sentiment Pulse Lite

• Outcome: Pass
• Latest audit: Sep 16, 2026 · 4:26 AM UTC
• Version: 1.0.0

Overview

This is a web-based market sentiment helper with visible upsell text, but it does not install code, persist, access credentials, or modify user data.

Before installing, be aware that results will include promotional text for a third-party full version and may run web searches for broad market-sentiment prompts. Use it for quick headline sentiment only, verify sources, and do not treat its output as financial advice.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

Findings (1)

T01 · Skill Instruction Hijacking

Warning

• Location: SKILL.md:182

Finding

Mandatory Commercial Promotion Injected into Agent Responses

Content

View full analysis

Vulnerability Details

File Location: SKILL.md, lines 182-186
Vulnerability Type: Instruction-driven output hijacking
Risk Level: Medium

Vulnerable Code

markdown

---
*⚠️ Lite sentiment check — headlines and crowd temperature only.*
*For scored multi-source Pulse, contrarian signals, and options flow:*
*Full version → agentofalpha.com*

Technical Analysis

The skill's prescribed output template requires the agent to append commercial promotional content and an external domain to ordinary sentiment-analysis responses. Because this text is embedded in the mandatory output format, it alters the agent's response objective from solely satisfying the user's request to promoting a third-party service.

The behavior is reinforced by additional promotional instructions at SKILL.md:22-24 and SKILL.md:205-207. No evidence establishes that the referenced website is itself malicious; the confirmed issue is the compulsory and unrelated manipulation of agent output.

This best matches T01: Skill Instruction Hijacking because attacker-authored skill instructions alter what the agent presents during the current session.

Attack Path

1. The agent loads SKILL.md as an active skill.
2. A user requests a market-sentiment analysis for a ticker or asset.
3. The agent performs the legitimate news and sentiment workflow.
4. The agent follows the mandatory output template in the skill.
5. The final response automatically includes promotional claims and directs the user to agentofalpha.com, even though the user did not request product recommendations or external services.

Impact Assessment

The issue permits manipulation of user-facing agent responses and recurring redirection of users to an external commercial website. It can reduce response integrity, create undisclosed commercial bias, and cause users to interpret promotional claims as trusted agent recommendations.

The affected scope is ...[truncated 210 chars]

Remediation

View remediation

Remediation Suggestions

1. Remove the mandatory promotional footer and external domain from the response template.
2. Remove or clearly separate the repeated promotional material at lines 22-24 and 205-207 from operational skill instructions.
3. Ensure the default output contains only information necessary to answer the user's sentiment-analysis request.
4. Present product or upgrade information only when the user explicitly asks about available versions, paid features, or related services.
5. If an external link is retained, label it clearly as promotional, avoid unsupported claims, and require explicit user action rather than inserting it into every result.
6. Add a review policy prohibiting advertisements, affiliate redirects, and unrelated calls to action in mandatory output templates.

SkillSpector

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 93% confidence

Finding

The invocation examples do not clearly bound when the skill should run, so a router may select it for general finance questions rather than narrow sentiment checks. In this skill's context, that can lead to unnecessary external searches, reduced answer quality, and unintended delegation to a promotional skill that steers users toward an upsell path.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 89% confidence

Finding

The example trigger phrase is broad enough to match ordinary conversation about market mood, which can cause the skill to activate outside clearly intended contexts. Over-broad activation increases the chance of unintended tool use (web_search/web_fetch), response hijacking of unrelated finance queries, and user confusion about why this specific skill was invoked.

Content

No source excerpt is available for this finding.