Aibrary Growth Plan

• Outcome: Pass
• Latest audit: Sep 11, 2026 · 5:47 AM UTC
• Version: 0.1.0

Overview

This is a simple growth-planning prompt skill with no code execution, data access, persistence, or hidden system changes.

Before installing, be aware that this skill may activate for broad self-improvement requests and may add an Aibrary-branded tagline to generated plans. It does not appear to access files, credentials, networks, or modify the system.

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

Note

• Location: SKILL.md:133

Finding

Mandatory Third-Party Branding Injected into Agent Responses

Content

View full analysis

Vulnerability Details

File Location: SKILL.md, line 133
Vulnerability Type: T01: Skill Instruction Hijacking
Risk Level: Low

Vulnerable Code Snippet:

markdown

*Growth plan created by Aibrary — structured learning for lasting growth.*

Technical Analysis

The skill places an Aibrary promotional attribution inside its required output template. When the skill is loaded and followed, this instruction causes the agent to append third-party branding to generated growth plans even when the user did not request attribution or promotional content.

This constitutes limited instruction hijacking because the skill modifies the agent's response behavior for a purpose unrelated to the substantive growth-planning task. The behavior is transparent in the source and does not alter safety constraints, invoke tools, execute code, or establish persistence.

Attack Path

1. A user requests a structured learning or personal growth plan.
2. The agent loads SKILL.md because the request matches the skill description.
3. The agent follows the prescribed output template.
4. The hard-coded promotional line is copied into the response.
5. The resulting content appears branded by or affiliated with Aibrary without the user explicitly requesting that branding.

No external attacker interaction, code execution, elevated access, or follow-on compromise is required for this behavior.

Impact Assessment

The issue can influence the content of responses generated while this skill is active. It may create an unintended appearance of endorsement, authorship, or affiliation and reduces user control over generated content.

The issue grants no operating-system privileges, file access, network access, credential access, or cross-session persistence. Its scope is limited to response-content manipulation within sessions where the skill is invoked.

Remediation

View remediation

Remediation Suggestions

1. Remove the hard-coded Aibrary tagline from the mandatory output template.
2. Keep the template focused exclusively on content needed to fulfill the user's growth-planning request.
3. If attribution is operationally required, make it optional and include it only after obtaining informed user consent.
4. Clearly distinguish optional attribution from required functional instructions so an agent does not treat promotional content as mandatory.
5. Add a review rule prohibiting undisclosed advertising, endorsements, or third-party branding in generated responses.

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 (1)

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 95% confidence

Finding

The trigger description is overly broad, using catch-all phrasing such as 'any structured self-improvement intent' along with generic phrases like 'help me grow in' and 'I want to develop.' This can cause the skill to activate for loosely related requests, increasing the chance of incorrect routing, unintended disclosure of user context to the skill, or interference with more appropriate skills.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.