Security audit

crypto-market-analyst

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent crypto market analysis skill that uses a disclosed Finskills API key and market-data endpoints, with no evidence of hidden execution, persistence, or data exfiltration.

Before installing, be comfortable providing a Finskills API key and allowing the agent to make market-data requests to Finskills. The skill produces crypto research and positioning signals, not personalized financial advice, and its documented workflow appears to rely on price, market, history, and macro proxy data rather than true on-chain metrics.

A.I.G

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

SkillSpector

Vulnerability Patterns

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
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The trigger phrase "What's the current state of the crypto market?" is broad, natural language that closely overlaps with ordinary user conversation. In agent systems that auto-route or invoke skills from trigger text, this can cause unintended activation of the skill in unrelated contexts, leading to prompt-scope confusion or unnecessary external API access. The crypto-analysis context makes this somewhat more plausible because users commonly ask broad market questions in exactly this wording.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The manifest description explicitly promises analysis using 'on-chain signals,' but the remainder of the file only specifies price, market cap, dominance, historical returns, volatility, and macro proxies from API endpoints. This is an active contradiction between the claimed analytical basis and the documented behavior of the skill, not merely an omitted implementation detail.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The activation section includes general triggers such as asking about crypto prices, market conditions, comparisons, or current risk/opportunity, without clear exclusion conditions or boundaries. These phrases are broad enough to match ordinary conversation about cryptocurrency and could cause unintended activation when the user is not specifically requesting this skill's full market-analysis workflow.
Content: No source excerpt is available for this finding.

Description-Behavior Mismatch

Low

Category: Not specified by scanner
Confidence: 79% confidence
Finding: The manifest description says the skill analyzes crypto using real-time price, on-chain signals, and macro context from the Finskills API, which broadly fits the skill’s purpose. However, the implementation guidance explicitly expands data retrieval to broader market and commodity endpoints such as /v1/market/summary and /v1/free/commodity/price/gold, which goes beyond the narrowly described crypto-focused inputs and introduces additional non-crypto market analysis behavior.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.