Security audit

Continuous User Research

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed user-research workflow with consent, redaction, and retention safeguards, and I found no hidden code or deceptive behavior.

Install this only for product research where you can obtain participant consent, restrict raw data access, set retention windows, and use least-privilege integration tokens. Be especially careful with screenshots, audio, Slack email lookup, and optional ticket creation, since those can expose participant identifiers or business data if configured too broadly.

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

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 (4)

Autonomous Decision Making

Medium

Category: Excessive Agency
Confidence: 80% confidence
Finding: Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
Content: Scanner excerpt · SKILL.md (reported line 29)May include surrounding context.
md
- continuous_user_research.retention.redacted_report_days --- Automatically run a lightweight diary study and turn it into weekly product signals and experiment-ready recommendations. ## When to use - The team must decide what to build, fix, or prioritize and needs behavior evidence over time, not one-off opinions.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The template explicitly solicits screenshots/photos and voice notes, which can easily contain personal data, account details, device information, background content, or other sensitive identifiers beyond what the participant intends to share. Although it includes a brief instruction to remove or blur identifiers, it does not clearly explain privacy risks for audio submissions, does not require minimization by default, and may normalize collecting richer data than necessary in a longitudinal research context.
Content: No source excerpt is available for this finding.

Missing User Warnings

Medium

Category: Not specified by scanner
Confidence: 84% confidence
Finding: This markdown recipe instructs operators to resolve Slack users by email and call users.lookupByEmail, which involves handling participant identifiers. While the file includes security controls, it does not explicitly warn users about the privacy implications of using email-based identity matching in Slack workflows.
Content: No source excerpt is available for this finding.

Vague Triggers

Low

Category: Not specified by scanner
Confidence: 81% confidence
Finding: This file is JSON, so vague-trigger checks apply. The title "Continuous User Research Study Brief" is descriptive but does not define any explicit trigger phrases, scope limitations, or exclusion conditions; if this schema metadata is used by a skill-discovery or activation layer, the broad wording could overlap with many generic research-related requests.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.