Hydration Habit Tuner

• Outcome: Pass
• Latest audit: Sep 16, 2026 · 8:45 AM UTC
• Version: 1.0.1

Overview

This is a document-only hydration habit coaching skill with clear health safety boundaries and no code, network, credential, or persistence behavior.

Reasonable to install for simple hydration habit planning. Do not use it as medical advice, and avoid sharing more health information than needed; users with fluid restrictions, kidney or heart conditions, pregnancy-related guidance, severe dehydration symptoms, or urgent symptoms should follow clinician or emergency guidance.

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
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (5)

Natural-Language Policy Violations

High

• Confidence: 98% confidence
• Finding: The verification checklist goes beyond metadata and enforces an English-only, ASCII-safe, no-CJK/no-RTL policy, which is a stronger exclusionary control affecting who can use or contribute to the skill. In a general wellness tool, this unjustified locale restriction can systematically block non-English users and scripts, creating accessibility and fairness risk even though it is not a traditional exploit path.

Credential Access

High

• Category: Privilege Escalation

Content

- **Secrets scan:** No API keys, tokens, passwords, or private identifiers present.
- **Executable scan:** No scripts (.sh, .py, .js, .rb), no package files, no build artifacts.
- **Network scan:** No API calls, no HTTP client usage, no webhook URLs, no remote endpoints.
- **Credential scan:** No credential instructions, no .env references, no auth flow.
- **Safety metadata:** document_only, promptOnly, runtime:none, execution:noExec, no_code_execution, no_network, no_credentials, requires_api:false.
- **Encoding:** English-only (ASCII-safe), no CJK, no RTL, no special Unicode.
- **File count:** 3 files (SKILL.md, ACCEPTANCE.md, skill.json), no temp/log/hidden files.

• Confidence: 60% confidence
• Finding: Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Anti-Refusal Statement

High

• Category: Anti-Refusal

Content

### Step 2: Identify Dry Spells

A dry spell is a part of the day when the user routinely goes a long time without drinking anything. Do not judge it. Treat it as a design problem.

Dry spell review:

• Confidence: 85% confidence
• Finding: Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Natural-Language Policy Violations

Medium

• Confidence: 92% confidence
• Finding: The acceptance criteria explicitly set the skill language to English, which creates an unnecessary language restriction that can exclude users without a documented safety or technical justification. In a health-adjacent habit coaching skill, this can reduce accessibility and equitable use, though it does not directly enable code execution, data theft, or system compromise.

Natural-Language Policy Violations

Low

• Confidence: 94% confidence
• Finding: The manifest sets "language": "en", which specifies a single language/locale with no indication that users can opt into another language or that the skill is intentionally region-specific. Under the stated policy, forcing a specific language without user opt-in is a natural-language policy concern.

Static analysis

No suspicious patterns detected.