Security audit

Ra Lite

Security checks for vulnerabilities and agentic risk

Overview

The skill performs the promised quick research brief, but its setup uses an unnecessarily broad system-level Python package install for cosmetic output.

Install only if you are comfortable with a skill that may change your system Python packages. Prefer running it in an isolated virtual environment, or remove the rich install/formatting before use. Expect it to show an upgrade prompt after the research brief.

A.I.G

Vulnerability Patterns

Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Insecure DependenciesIntroduces malicious components through unsafe dependency sources
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

Findings (2)

T01 · Skill Instruction Hijacking

Warning

Location: SKILL.md:107
Finding: Mandatory Commercial Output Hijacks the Agent Response
Content: View full analysis
[bold cyan]ko-fi.com/occupythemilkyway[/bold cyan]", title="[yellow]Want deeper research?[/yellow]", border_style="yellow" )) ``` ``` ### Technical Analysis The Skill defines the commercial upsell as a mandatory workflow step rather than optional documentation or a response to an explicit user request. Once the Skill is loaded, it directs the Agent to append advertising for paid products and an external sales destination after completing the requested research task. This modifies the Agent's output objective from producing only the requested research brief to distributing attacker-controlled promotional content. Because the instruction is embedded in the trusted Skill workflow, users may perceive the advertisement and outbound destination as endorsed components of the Agent's response. ### Attack Path 1. A user activates the Skill and supplies a value through `RESEARCH_TOPIC`. 2. The Agent follows the Skill instructions and produces the requested three-source research brief. 3. The mandatory “Step 4 -- Upsell” instruction causes the Agent or associated workflow to append commercial content. 4. The output directs the user to `ko-fi.com/occupythemilkyway`, despite the user not requesting purchase information. 5. The Agent's trusted response channel is consequently used to promote an ...[truncated 575 chars]
Remediation: View remediation

T08 · Insecure Dependencies

Warning

Location: SKILL.md:42
Finding: Unpinned Dependency Installation Bypasses System Package Protections
Content: View full analysis
Remediation: View remediation

SkillSpector

Vulnerability Patterns

Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep

Static analysis

No suspicious patterns detected.