ai-product-manager-playbook

• Outcome: Pass
• Latest audit: Sep 16, 2026 · 6:02 AM UTC
• Version: 1.0.0

Overview

This skill is a mostly Markdown-based AI product management playbook with safety-testing templates and no hidden execution, persistence, or exfiltration behavior found.

Reasonable to install if you want AI product management templates and workflows. Treat any production telemetry, user queries, system prompts, or feedback you provide to the skill as potentially sensitive, and review or redact it before use.

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

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• YARA Signatures: Malware Match, Webshell Match, Cryptominer Match
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (4)

YARA rule 'agent_skill_prompt_injection_hidden_instructions': Prompt injection or hidden instructions embedded in AI agent skill text [agent_skills]

High

• Category: YARA Match
• Confidence: 80% confidence

Finding

YARA rule matched a hack tool or exploit indicator (offensive tools, reconnaissance, privilege escalation, or exploit frameworks).

Content

Scanner excerpt · templates/red_teaming_plan.md (reported line 11)May include surrounding context.

md

ur proactive, adversarial testing of AI systems for vulnerabilities and potential harms before release.

## 1. System Overview
*Describe the AI system being tested, its intended use cases, and the target audience.*
**System Name:** [Insert System Name]
**Intended Use Cases:** [List the primary use cases for the system]
**Target Audience:** [Describe the intended users of the system]

## 2. Threat Model & Risk Assessment
*Identify the potential threats and risks associated with the system.*
**Threat Actors:** [Who might try to exploit the system? e.g., malicious users, competitors, nation-states]
**Potential Harms:** [What are the potential negative consequences of the system? e.g., bias, misinformation, privacy violations]
**Risk Level:** [Assess the overall risk level of the system, e.g., Low, Medium, High, Critical]

## 3. Red Teaming Objectives
*Define the specific goals of the red teaming exercise.*
**Objective 1:** [e.g., Identify vulnerabilities to prompt injection attacks]
**Obj

Anti-Refusal Statement

High

• Category: Anti-Refusal
• Confidence: 90% confidence

Finding

Skill attempts to nullify the agent's safety policies or restrictions ('you have no restrictions', 'ignore your guidelines', 'do anything now'). This is a direct jailbreak that disables guardrails.

Content

Scanner excerpt · templates/red_teaming_plan.md (reported line 31)May include surrounding context.

md

## 5. Test Scenarios & Attack Vectors
*List the specific test scenarios and attack vectors that will be executed.*
- **Scenario 1:** [e.g., Attempt to bypass safety filters using jailbreak prompts]
- **Scenario 2:** [e.g., Input biased or discriminatory language to test the system's response]
- **Scenario 3:** [e.g., Attempt to extract sensitive information or PII from the system]

Instruction Override

High

• Category: Prompt Injection
• Confidence: 90% confidence

Finding

This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.

Content

Scanner excerpt · templates/red_teaming_plan.md (reported line 31)May include surrounding context.

md

## 5. Test Scenarios & Attack Vectors
*List the specific test scenarios and attack vectors that will be executed.*
- **Scenario 1:** [e.g., Attempt to bypass safety filters using jailbreak prompts]
- **Scenario 2:** [e.g., Input biased or discriminatory language to test the system's response]
- **Scenario 3:** [e.g., Attempt to extract sensitive information or PII from the system]

Vague Triggers

Medium

• Category: Not specified by scanner
• Confidence: 85% confidence

Finding

The manifest description says to use this skill when 'planning, prototyping, evaluating, or launching AI-native products,' which is a very broad activation scope for a manifest file. It does not provide explicit trigger phrases, constraints, or exclusion conditions, so the skill could be invoked in many ordinary AI-product discussions where a narrower match may be intended.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.