Security audit

Competitive Analysis

Security checks across malware telemetry and agentic risk

Overview

The skill is a competitor-analysis workflow that includes plaintext third-party login credentials and instructions to reuse authenticated sessions, so it should be reviewed carefully before installation.

Install only if you have explicit authorization to access the named third-party accounts and collect this information. Remove the plaintext credentials, rotate any exposed passwords, avoid persistent cookie files, and require interactive user-approved login through a secure secret mechanism before using the skill.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (11)

Context-Inappropriate Capability

Medium

Confidence: 94% confidence
Finding: The skill explicitly directs use of a separate file containing login credentials and describes cookie management for authenticated competitor accounts. That expands the capability from analysis into handling and reusing sensitive authentication material, which creates account-compromise, unauthorized-access, and secret-exposure risk if the workspace, logs, or artifacts are accessed by others.

Context-Inappropriate Capability

Medium

Confidence: 96% confidence
Finding: The documentation instructs persistent storage of authenticated cookies on disk, which are bearer tokens that may grant direct access without reauthentication. Persisting them locally increases the chance of session theft, replay, lateral misuse by other tools/users, and accidental inclusion in backups or reports.

Context-Inappropriate Capability

High

Confidence: 99% confidence
Finding: The file contains plaintext third-party account credentials, login URLs, operational post-login steps, and cookie storage guidance that enable reuse of authenticated access. In the context of a competitor-analysis skill, this is especially dangerous because it turns a research workflow into a repository for unauthorized account access and session persistence against external services.

Vague Triggers

Medium

Confidence: 78% confidence
Finding: The trigger phrases are broad enough to match ordinary product-research requests, which raises the chance the skill activates in contexts where users did not intend authenticated browser automation against third-party systems. In this skill, that matters because activation can lead to credential use, protected-site access, and sensitive data collection.

Missing User Warnings

High

Confidence: 97% confidence
Finding: The skill tells the agent to access login credentials from a sensitive reference file without a prominent user-facing warning or consent boundary for handling secrets. That normalizes silent use of third-party credentials and increases the risk of unauthorized access, accidental disclosure, and misuse beyond the user's expectations.

Missing User Warnings

Medium

Confidence: 95% confidence
Finding: The skill describes local retention of session cookies but does not warn that these artifacts are equivalent to live authentication tokens. Without clear handling guidance, operators may store, copy, or share them insecurely, enabling account takeover or unintended continued access.

Missing User Warnings

High

Confidence: 97% confidence
Finding: This markdown documents reusable credentials and prescribes storing cookies on disk by platform and role, which creates a clear path for session theft, replay, and lateral misuse. The lack of risk controls or warnings is not the core issue; the dangerous part is the normalization of insecure secret and session handling in operational documentation.

Ssd 3

High

Confidence: 98% confidence
Finding: Directing the agent to read stored credentials from a file in order to access competitor systems is a clear sensitive-data misuse pattern. It operationalizes third-party account access from embedded secret material, increasing the risk of unauthorized access, secret sprawl, and abuse if the environment or repository is compromised.

Ssd 3

High

Confidence: 98% confidence
Finding: The workflow explicitly instructs reading account credentials and logging into competitor platforms as part of normal execution. That turns the skill into an automation path for authenticated access to third-party systems, which is highly sensitive and can violate authorization boundaries, contractual terms, or data-handling expectations.

Ssd 4

High

Confidence: 97% confidence
Finding: The document lays out an end-to-end process to obtain access, maintain sessions, traverse protected areas, and systematically harvest internal product details from competitor systems. In context, this materially increases danger because the skill is purpose-built for deep authenticated reconnaissance, making unauthorized collection of non-public business information scalable and repeatable.

Ssd 3

High

Confidence: 99% confidence
Finding: The file systematically organizes sensitive login credentials, system entry points, account context, and reusable session artifact practices for multiple third-party platforms. That structure materially lowers the barrier to unauthorized access, credential abuse, and durable session hijacking, making the skill context more dangerous rather than less because competitor analysis does not justify embedding external account access details.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.