Pollen rapporten

• Outcome: Pass
• Latest audit: Sep 11, 2026 · 7:22 PM UTC
• Version: 1.0.0

Overview

This skill does what it claims: it fetches Swedish pollen forecasts from a disclosed public API, with no hidden persistence, credential access, or file mutation.

Before installing, understand that forecast lookups contact api.pollenrapporten.se and depend on that service. Use the grass-pollen alert output as general allergy guidance only, not as personalized medical advice.

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

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands

Findings (9)

Lp3

Medium

• Category: MCP Least Privilege
• Confidence: 92% confidence
• Finding: The skill invokes a Python script that fetches data from an external API, but the manifest does not declare any tool scope or permissions for network use. Missing capability declarations weaken governance and review controls, making it easier for a skill to perform external communication without explicit approval or user visibility.

External Transmission

Medium

• Category: Data Exfiltration

Content

## Data Sources

- **API**: https://api.pollenrapporten.se/v1/
- **Source**: Swedish Museum of Natural History (Naturhistoriska riksmuseet)
- **Coverage**: 24 regions across Sweden
- **Update frequency**: Daily during pollen season

• Confidence: 74% confidence
• Finding: The skill sends user-supplied location queries to an external third-party API, which is an external data transmission event. In context, this is expected functionality for a forecast skill, but it still creates privacy and supply-chain risk because user input and fetched content leave the local environment and depend on a remote service.

Natural-Language Policy Violations

Medium

• Confidence: 94% confidence
• Finding: The note says 'You must send in Swedish, and also in English,' which imposes a language behavior unconditionally. This is a natural-language policy concern because it forces a specific locale/language choice rather than offering the user an option or documenting a justified region-specific exception for output language.

External Transmission

Medium

• Category: Data Exfiltration

Content

import urllib.request
from pathlib import Path

API_BASE = "https://api.pollenrapporten.se/v1"
SKILL_DIR = Path(__file__).parent.parent
REGIONS_FILE = SKILL_DIR / "references" / "regions.json"

• Confidence: 60% confidence
• Finding: Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Intent-Code Divergence

Low

• Confidence: 80% confidence
• Finding: Line L134 states 'You must send in Swedish, and also in English,' which implies bilingual output behavior. Elsewhere in the file, the documented implementation only fetches Swedish forecasts from the API and formats them; no translation step or dual-language rendering is described in the examples or workflow.

Natural-Language Policy Violations

Low

• Confidence: 81% confidence
• Finding: This JSON file is entirely built around Swedish region names and mappings, including accented Swedish place names and English-to-Swedish aliases. Because SQP-3 applies to all file types, this can be flagged as a locale-specific constraint that is encoded without any accompanying natural-language justification or user opt-in in the file itself.

Intent-Code Divergence

Low

• Confidence: 77% confidence
• Finding: The top-level docstring frames the script as a simple forecast fetcher, while the implementation supports an extra '--alert' mode that produces tailored warning and medication/behavior recommendations. This is more than an omitted implementation detail because it changes the apparent intent from data retrieval to advisory alerting.

Natural-Language Policy Violations

Low

• Confidence: 86% confidence
• Finding: The script hardcodes bilingual and Swedish-specific output strings such as level names and forecast text labels, while the usage and other messages are in English. This imposes a locale choice in the user-facing output without any opt-in or configuration mechanism.

Context-Inappropriate Capability

Low

• Confidence: 88% confidence
• Finding: The manifest describes a Sweden pollen forecast skill that reports pollen levels and severity. The code adds a special-purpose capability keyed to a specific user's grass allergy and later emits behavioral and medication recommendations, which is not necessary to fetch or present forecast data and is not declared in the skill description.

Static analysis

No suspicious patterns detected.